BIOPHYSICAL PROPERTIES OF VEGETATION
Biomass - above-ground
Biomass - below-ground
Leaf area index (LAI)
Net primary productivity (NPP)
Net ecosystem productivity (NEP)
Necromass
Plant tissue nitrogen and phosphorus content
Roughness - surface
Spectral vegetation greenness index
Stomatal conductance - maximum
Vegetation structure
BIOGEOCHEMISTRY
Biogeochemical transport from land to oceans
Biomass - peak leaf of nitrogen-fixing plants
Dissolved organic C, N, and P in water (rivers and lakes)
Dry deposition of nitrate and sulphate
Emissions of CO2, NOX and SOX from combustion of fossil fuels
Fertiliser use N and P
Peak leaf biomass of nitrogen-fixing plants
Rainfall chemistry
Volcanic sulphate aerosols
LAND COVER/LAND USE AND DISTURBANCE
Fire area
Land cover
Land use
SOIL PROPERTIES
Rooting depth - 95%
Soil cation exchange capacity
Soil moisture
Soil total carbon
Soil total nitrogen
Soil phosphorus - available
Soil phosphorus - total
Soil bulk density
Soil particle size distribution
Soil pH
Soil surface state
HYDROLOGY
Biogeochemical transport from land to oceans
Evapotranspiration
Ground water storage fluxes
Precipitation - accumulated (solid and liquid)
Relative humidity (atmospheric water content near the surface)
Surface water flow - discharge
Surface water storage fluxes
CRYOSPHERIC PROPERTIES
Firn temperature (ice sheets, ice caps, glaciers)
Glaciers and ice caps
Ice sheet mass balance
Ice sheet geometry
Lake and river freeze-up and break-up (timing)
Permafrost - active layer
Permafrost - thermal state
Sea ice concentration/extent
Sea ice motion
Sea ice thickness
Snow cover area
Snow depth
Snow water equivalent (SWE)
RADIATION (AND RELATED VARIABLES)
Aerosols
Albedo
Cloud cover
Radiation - fraction of photosynthetically active radiation (FPAR)
Radiation - incoming short-wave
Radiation - outgoing long-wave
Radiation - reflected short-wave
Temperature - air
TRACE GASES
Carbon dioxide flux (CO2)
Methane flux (CH4)
ANCILLARY VARIABLES
Topography
Wind velocity
AEROSOLS
Units of measure:
Atmospheric transmissivity.
Users:
Hydrologists, climatologists, ecologists, agronomists, remote-sensors.
Rationale:
Aerosols are not needed by terrestrial scientists in their own right, but for the calculation of incoming radiation. They are also useful indices of biomass burning and wind erosion.
Frequency of measurement:
Hourly measurements at centres equipped with tracking sun photometers; 10-daily climatologies from satellite remote-sensing.
Spatial resolution:
100 km.
Accuracy/precision required:
± 10%.
R and D needed:
Development of robust all-weather infra-red sensors.
Present status:
Not routinely collected except by satellites such as AVHRR.
Measurement method:
Tiers 1-2: spectroradiometric measurements;
Tier 5: spectral radiance.
Associated measurements:
Incoming short wave, outgoing short wave.
Action required:
Promote the establishment of a sun photometer network.
BIOGEOCHEMICAL TRANSPORT FROM LAND TO OCEANS
Units of measure:
mg/l (fluxes should be calculated from raw data in metric tonnes per year), particle shape and size distribution (mm).
Users:
Earth system modellers; carbon cycle modellers; natural resource assessments; climate modellers (eventually); those concerned with detection of change and general health of the continental land masses.
Rationale:
Biogeochemistry focuses on the carbon nitrogen, phosphorus and sulphur cycles, and underlies all primary production and trace gas emission models. The impact on the climate is via the emissions of radiatively-active trace gases which occur as a result of biogeochemical cycling (CO2, CH4, N2O, ozone precursors such as NO, CO and NMVOC, and aerosols) and via the land surface characteristics such as biomass and leaf area which are constrained by biogeochemical considerations. Since biogeochemical cycling is strongly influenced by climate, this constitutes one of the major avenues for both impacts and feedbacks.
Ocean systems are overall net stores of carbon and other biogeochemicals through burial. The system is kept "topped-up" by inputs from continental areas. In the case of carbon, estimates of around 1 Gt/a (1% terrestrial NPP) are often quoted. Shelf seas are the initial recipients of the nutrient fluxes, but less than half is accumulated in shallow seas. The majority is dispersed through the various food chains into the open ocean.
For calculation of the fluxes between the land, freshwater hydrosphere and oceans, stream flow measurements are critical.
Frequency of measurement:
Daily measurements of dissolved and particulate organic carbon, nitrogen and phosphorus, nitrate, pH, phosphate, dissolved oxygen, water temperature, suspended and bedload sediment as concentrations scaled to volume by discharge data. Particle size and shape distribution seasonal.
Spatial resolution:
Lowest practical gauging point on the world's major rivers, say average of 20 per continent. To the extent possible co-locate discharge and water quality stations close to each other.
Accuracy/precision required:
10% of true value.
R and D needed:
Further development of automated techniques of analysis. Associate measurements with basin characteristics to fill unmonitored basins to complete continental budgets. Use of remote-sensing for determination of sediment load.
Present status:
Data already collected on majority of rivers and reasonable body of research has established average values (guessed precision ± 25%). GEMS water has a number of locations that are currently collecting the appropriate information, but they lack comprehensive country participation.
Measurement method:
All tiers direct measurement of all chemical variables, automatic gauging stations for flow.
Associated measurements:
Runoff at monthly time-steps and water balance variables representing large land areas. Elements of the land and ocean carbon and nutrient cycle.
Action required:
Bring the appropriate existing GEMS water stations into the network and strengthen these in cooperation with Land Ocean Interactions in the Coastal Zone (LOICZ) and FRIEND;
Upgrade and/or rehabilitate existing stations.
BIOMASS - Above-ground
Units of measure:
g/m2.
Users:
Ecosystem and carbon modellers, decision-makers concerned with international measures to mitigate atmospheric carbon increases.
Rationale:
Biomass is a key variable in annual and long-term changes in the global terrestrial carbon cycle, and in modelling carbon uptake and redistribution within the ecosystem. Of most interest is the live woody biomass. It mediates and buffers atmospheric carbon concentrations and thus its dynamics must be understood if annual spatial variations in atmospheric CO2 are to be related to spatial weather variables. It is also a key variable (along with carbon emissions and carbon sequestration rates) in defining carbon status and flux in a given geopolitical unit for assessment of carbon taxes, and similar international atmospheric CO2 mitigation measures.
Frequency of measurement:
Once every 5 years.
Spatial resolution:
0.01 to 1 km2.
Accuracy/precision required:
± 10%.
R and D needed:
Research is required to determine the feasibility of estimating live biomass from satellite data; radar measurements presently offer the best prospect. Imaging laser sensor technology is also promising but no space-borne missions are planned.
Present status:
Data available are very coarse-grained in developed countries, almost non-existent in less-developed countries; resulting estimates vary widely.
Measurement method:
Tiers 1-3: destructive sampling, allometric equations;
Tier 5: inferred from spectral radiance measurements (non-woody types, experimental stage).
Associated measurements:
Below-ground biomass, NEP, necromass, vegetation structure, vegetation roughness.
Action required:
Currently, biomass is determined from labour-intensive techniques involving sampling, drying and weighing all vegetation on a given unit area, including whole tree and root system harvesting. Beyond these techniques, it is uncertain how we can more efficiently get biomass data for the modelling efforts within the next 5 years;
Encourage R and D on biomass estimation from now highly-experimental satellite techniques.
BIOMASS - Below-ground
Units of measure:
g/m2 or kg/ha.
Users:
Terrestrial ecologists, ecosystem modellers, decision-makers (e.g., for calculating carbon taxes, etc.).
Rationale:
Above- and below-ground biomass are key variables in annual and long-term changes in the global terrestrial carbon cycle, and in modelling carbon uptake and redistribution within the ecosystem. Experimental evidence indicates that changes in atmospheric chemistry and climate will change carbon allocation patterns between the above-ground and below-ground pools, hence, they will not change in parallel as a response to global environmental change. The currently large but undetected terrestrial carbon sink proposed by climate modellers and oceanographers may well be found eventually in below-ground biomass.
Frequency of measurement:
After IOS, annually to every 5 or 10 years except for centre and land experiment analyses, the needs for which then may dictate seasonal or monthly measurements.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
+ 10%.
R and D needed:
Currently, there is neither sufficient theory on carbon allocation in plants under chronic environmental change to predict below-ground biomass increments from other variables (i.e., temporal patterns) nor is there enough data to define current spatial patterns and their relationship to other spatial variables. In addition, the only measurement technique is the labour-intensive removal, drying and weighing of below-ground plant parts. Measurement techniques need to be developed to measure current distributions and distribution changes through time.
Present status:
Below-ground biomass is known from only a few sites around the world, most of them associated with one-time experiments or with research stations. This is wholly inadequate for global change purposes.
Measurement method:
Tiers 1-3: sampling; video cameras (fine roots; experimental).
Associated measurements:
Above-ground biomass is most closely associated, as the two are required to define biomass distributions and increments, while neither alone is of much value in terrestrial carbon cycle understanding or prediction.
Action required:
Currently below-ground biomass is determined from labour-intensive techniques involving extraction, drying and weighing all roots on a given unit area. Beyond these techniques, it is uncertain how we can more efficiently obtain biomass data for the modelling efforts within the next 5 years;
Encourage R and D on biomass estimation from now highly-experimental radar imagery techniques.
BIOMASS - PEAK LEAF OF NITROGEN-FIXING PLANTS
Units of measure:
g/m2.
Users:
BCG and ecological modellers, Greenhouse Gas (GHG) modellers.
Rationale:
Nitrogen fixation by microbes associated with certain species of plants is one of the major sources of nitrogen inputs to terrestrial ecosystems and some forms of agriculture. The magnitude of the input varies according to which crop is planted, and may be very sensitive to the atmospheric CO2 concentration in natural ecosystems. The direct measurement of biological nitrogen fixation (BNF) is difficult and costly. Indirect indices, such as the reduction of acetylene by soils, are simpler, but still tedious. An approximate index can be obtained from the fraction of the total plant biomass which is made up by plant species known to have associated nitrogen fixing organisms. When considered in conjunction with the leaf N content (required elsewhere) and especially its isotopic content (not required), quantitative estimates of BNF are possible.
Frequency of measurement:
Per crop of BNF plants in agricultural systems, once a decade in natural systems.
Spatial resolution:
A representative sample per site.
Accuracy/precision required:
± 10%.
R and D needed:
Compilation of BNF rates for different functional types of BNF-associated plants from the research literature.
Present status:
No existing observing system, large literature of measurements.
Measurement method:
Requires knowledge of leaf area index (see leaf area index variable sheet) and species composition. Species composition is obtained through field observations (Tiers 1-4) or through remote sensing in combination with limited ground sampling (Tier 5).
Associated measurements:
Leaf N, leaf area index, cover type.
Action required:
Literature compilation.
CARBON DIOXIDE FLUX (CO2)
Units of measure:
ppmv.
Users:
Climate change and ecosystem modellers and researchers. Science policy advisors nationally and globally, e.g., IPCC.
Rationale:
Precise observations of past, present and future concentrations of atmospheric CO2 are critical in attempts to model and understand the global carbon cycle and possible CO2-induced climate change. Between 1980 and 1989, atmospheric CO2 increased on average approximately 1.5 ppmv (or 3.2 ±.2 GtC/yr) per annum. The rate of increase slowed significantly between 1991 and 1993, but increased subsequently in 1994. Consistent long-term observations are necessary to determine trends and identify causes of change. Atmospheric concentration represents the net effect of a variety of sources and sinks. An unidentified terrestrial sink(s) of 1.4 GtC/yr contrasts with tropical land use sources of 1.6 ± 1.0 GtC and northern forest sink of 0.5 ± 0.5 GtC/yr. Refining these estimates and monitoring change (e.g., the suggested shift from sink to source in tundra over the last decade) are critical. Monitoring of atmospheric CO2, land use change, and net CO2 flux from particular terrestrial systems, are the main requirements.
Frequency of measurement:
Continuous or monthly atmospheric concentration;
Cumulative or weekly/monthly 24 hr samples at tier 3 stations;
Continuous measures over short-term seasonal phases at tier 1 facilities and tier 2 centres.
Spatial resolution:
Approximately 100 land-based atmospheric samples sites;
Metre plots or km transects for representative tier 3 stations;
Intensive profile measurements over km fetch on tier 1 facilities and tier 2 centres.
Accuracy/precision required:
± 5% accuracy, but quantification of spatial variation is more important than the analytical error.
R and D needed:
Among the many requirements in this critical subject are: development of satellite and airborne sensors; development of long-path detection instruments for ground-based, landscape scale flux measurements; extensive field measurements with experimental treatments to assess, and then monitor, net ecosystem flux; and medium-term research to establish techniques to distinguish and monitor root and heterotrophic soil respiration.
Present status:
Atmospheric concentration monitoring is reasonably well established through the NOAA/Climate Monitoring and Diagnostic Laboratory (CMDL) air sampling network. Field-based observations are a completely sporadic and poorly-distributed research activity with lack of representation across biomes and agro-ecological zones. As a result, estimation sink/source relationships and changes have very high errors. A number of intensive, short-term, net ecosystem flux studies are making significant progress but technology is limiting. WMO's Global Atmosphere Watch (GAW) coordinates a network of sites to measure CO2 around the globe.
Measurement method:
Tier 1-2: flask samples and laboratory measurement; aircraft CO2 concentration profile measurements across horizontal and vertical ranges (experimental).
Associated measurements:
Quantitative definition of land cover and its derivative, land use, are the important basis for extrapolation from spot measurements. Systematic observation of climate, vegetation and soil characteristics are essential for calculations, comparisons and extrapolation.
Action required:
Production of guidelines for field observation and establishment of a network of replicated sites covering main biomes and agroecosystems (tier 2 centres and tier 3 stations).
CLOUD COVER
Units of measure:
Fractional cover; by type.
Users:
Hydrologists, climatologists, remote-sensors, ecologists.
Rationale:
Cloud cover is not needed in its own right, but as an intermediate variable for the calculation of incoming radiation. It is therefore not needed at the station and centre levels, but is needed at the level of global fields.
Frequency of measurement:
Hourly means from stations equipped with short-wave radiometers; daily, accumulated 10-daily for satellite products.
Spatial resolution:
10-50 km.
Accuracy/precision required:
± 10% absolute.
R and D needed:
Ongoing research on satellite-derived cloud cover products.
Present status:
Some existing cloud climatologies of inadequate spatial and temporal resolution. Work is in progress developing satellite-based cloud climatologies.
Measurement method:
Tiers 1-2: kilometers (sky cameras);
Tier 3: ocular estimates;
Tier 5: spectral radiance.
Associated measurements:
Incoming short wave, aerosols.
Action required:
Ensure the continuation of cloud climatology products.
DISSOLVED ORGANIC C, N, AND P IN WATER (Rivers and Lakes)
Units of measure:
mg NO3/litre, (mg Dissolved Organic N/litre, mg Particulate N/litre, mg Dissolved Organic C/litre, mg Particulate C/litre, mg PO4/litre, (mg Particulate P/litre, mg Dissolved Organic P/litre.
Users:
Biogeochemical modellers.
Rationale:
Phosphorus and nitrogen are the two nutrients most closely associated with primary productivity in freshwaters, including its extreme form, the eutrophication of rivers and lakes as a result of agricultural and urban pollution. Leaching of P and N into ground waters and then into rivers, lakes and ultimately the ocean, is an important avenue of N and P loss from terrestrial ecosystems, and a measure of their 'leakyness', often associated with anthropogenic disturbance. The level of nitrate in drinking water is a human and animal health issue.
The separation into dissolved inorganic, dissolved organic and particulate forms is necessary to capture all the major forms and to be able to attribute fluxes to processes. Dissolved organic forms are generally from relatively undisturbed natural ecosystems, inorganic forms are often the result of disturbance, excessive fertilisation or pollution, and particulate forms are associated with soil erosion.
Frequency of measurement:
Linked to the flow regime of rivers and the rate of turnover in water bodies; not less than monthly.
Spatial resolution:
One representative sample per site.
Accuracy/precision required:
5%.
R and D needed:
None.
Present status:
The measurements are widely made, but not globally standardised or observed.
Measurement method:
Tiers 1 to 3: direct measurement.
Associated measurements:
Flow in rivers.
Action required:
TBD.
DRY DEPOSITION OF NITRATE AND SULPHATE
Units of measure:
mg N/m2/month, mg S/m2/month.
Users:
BGC and ecological modellers, GHG modellers.
Rationale:
In arid regions, dry deposition of nitrate and sulphate is substantially greater than wet deposition, and forms a major biogeochemical input of N and S, with significant consequences to ecosystem functioning. Direct measurement methodology is not yet widely applicable. The proposal is to use simple passive diffusion collectors, which consists of a small capsule containing a filter and a membrane on which sulphate and nitrate are trapped. The captured N and S are washed off the membrane after a known exposure period and analysed. This is an index of the atmospheric content of reactive N and S, integrated over time, and the dry deposition rate is then modelled.
Frequency of measurement:
Weekly during episodes of high atmospheric content, monthly at other times.
Spatial resolution:
One sampler per site.
Accuracy/precision required:
10%.
R and D needed:
Models for dry deposition for all parts of the world.
Present status:
Applied in a research context (e.g., IGBP-International Global Atmospheric Chemistry Programme (IGAC) Deposition of Biogeochemically Important Trace Species (DEBITS) programme); incomplete global coverage and no planned continuity.
Measurement method:
Tiers 1-4: field sampling with standard filters;
Tier 5: modelling using atmospheric models (typically at the mesoscale level).
Associated measurements:
Rainfall chemistry, surface wind velocity, LAI, vegetation structure.
Action required:
Coordinate with IGBP-IGAC DEBITS.
EMISSIONS OF CO2, NOX, AND SOX FROM COMBUSTION OF FOSSIL FUELS
Units of measure:
Gg/country/year; resolved at sub-national scale where possible.
Users:
Climate and GHG modellers.
Rationale:
These values are required from Parties to the Framework Convention on Climate Change. The emissions of CO2 are a major forcing of the global atmosphere leading to global warming. SOX and NOX are significant sources of these elements leading to the acidification and nutrient forcing of terrestrial and freshwater ecosystems. SOX ultimately forms sulphate aerosols, important in the global radiation budget.
Frequency of measurement:
Every three years.
Spatial resolution:
National territory; resolved to 0.5° grid squares if possible.
Accuracy/precision required:
· TBD.
R and D needed:
None.
Present status:
Submitted to the FCCC Secretariat by Annex 1 countries of the FCCC.
Measurement method:
Not applicable (N/A).
Associated measurements:
CO2 content of atmosphere and its isotopic composition.
Action required:
Link to the FCCC Secretariat database.
EVAPOTRANSPIRATION
Units of measure:
kg of H2O/m2/day.
Users:
Evapotranspiration (ET) is a critical hydrological link between the Earth's surface and the atmosphere. It is therefore important for issues involving many aspects of climate, climate change, and ecosystem response. Evapotranspiration information is needed by climate modellers; climatologists; vegetation ecologists and physiologists; ecosystem modellers; hydrological and water resource specialists; and regional planners.
Rationale:
Evapotranspiration is the process responsible for the transfer of moisture from soil and vegetated surfaces to the atmosphere. Changes in evapotranspiration are likely to have large impacts on terrestrial vegetation, since the distribution and abundance of plant communities are controlled to a large extent by the quantity and seasonality of moisture. If changes in the water balance are significant, major shifts in vegetation patterns and condition are a likely result of climate change. Equally, changes in evapotranspiration are likely to impact atmospheric composition of greenhouse gases, and climate, as the hydrological cycle increases in intensity with warming. At the same time, increased atmospheric concentrations of CO2 are expected to decrease the rate at which leaves transpire by decreasing the frequency of stomata and the size of stomatal openings. It should be emphasized that actual evapotranspiration is the required measurement (not potential).
Frequency of measurement:
Daily or diurnal (4 × daily) measurements of potential evapotranspiration are adequate for most modelling and research applications. Higher frequencies, up to continuous measurement, are required for certain physiological research objectives.
Spatial resolution:
As with precipitation, a distinction is made under most headings between point and gridded values.
Point: 10-100 m2;
Gridded: 100-1 000 km2.
Accuracy/precision required:
Accuracy ± 20%.
R and D needed:
ET is an important parameter for the terrestrial hydrologic system, but difficult to measure because it is highly variable, both spatially and temporally. Many different methods exist to determine ET, but each is subject to fairly large biases and errors. It is still highly debatable how to aggregate ET measurements spatially. Evapotranspiration can be estimated by satellites from differences between leaf temperature and air temperature. Considering the inability of satellite optical imagery to accurately measure leaf or air temperature, this or unrelated and yet-to-be-tested radar imaging techniques, both require considerably more development to permit the regular ET estimates.
Present status:
Currently, potential evapotranspiration is measured at several thousand irregularly distributed sites worldwide from pan evaporation, or it is estimated from measurements of meteorological variables such as dry bulb and wet bulb temperature, net radiation, and atmospheric humidity. These approaches are too labour-intensive and indirect, respectively, to be adequate for a global programme. There now exist no practical means of measuring actual evapotranspiration over large areas. Eddy correlation offers a means of measuring actual evapotranspiration at a limited number of locations (selected in tiers 1-3). Most ET data will thus be derived estimates. At present, there is no consistent approach and no broad-based global data sets are available.
Measurement method:
Tiers 1-3: lysimeters, eddy correlation methods;
Tier 5: estimated using models driven by remote sensing and meteorological measurements.
Associated measurements:
Actual evapotranspiration is that amount of potential evapotranspiration permitted by the available soil moisture; humidity and wind are other variables which control evapotranspiration. In addition, this is one of a suite of weather variables often measured in the same place: solar radiation, air and soil temperature, evapotranspiration, precipitation, vertically structured wind speed and relative humidity.
Action required:
Despite its fundamental importance to the concept of the surface water balance, it is recommended that no spatially aggregated measure of ET be included for the IOS because of the extreme difficulty involved in producing a reliable global gridded data set. However, the IOS should include any ET measurements made at tier 1 facilities and tier 2 centres. All tier 3 sites should be upgraded to have automated temperature, humidity, wind and radiation measurements. It is important to ensure that the data collected presently (tier 1-3 sites) are archived and available;
For gridded products, there may be considerable potential value in the estimates of evapotranspiration derived by Soil-Vegetation-Atmosphere Transfer (SVAT) models, in particular those models contained within GCMs run to reanalyse past global data sets. It is understood that, at present, such estimates are not available and are possibly not even archived. However, together with remote-sensed data, they could provide gridded estimates of value to GCOS and GTOS;
The results from the WMO effort to correlate different ET methodologies as well as the effort by IGBP/BAHC (continuous flux network) should be available before a final decision is made as to what to achieve on a long-term basis. One possibility is to accept evapotranspiration time series that are consistently measured at a single site, and to include information about the nature of the ET measurement into metadata for one station.
FERTILIZER USE N and P
Units of measure:
Mg/country/year; resolved at a sub-national scale where possible.
Users:
BGC and ecological modellers, GHG modellers.
Rationale:
Fertiliser use is a major determinant of agricultural production. It is also a significant source of freshwater pollution, and an index of agricultural intensification. Nitrogenous fertiliser use is the major detriment of emissions of N2O from soils. The data on fertiliser manufacture and trade are generally available from international trade statistics and national industrial statistics. Use can be estimated from manufacture, plus net trade, plus net changes in stockpiles.
Frequency of measurement:
Annual.
Spatial resolution:
National territory; resolved if possible at subnational scale.
Accuracy/precision required:
10%.
R and D needed:
None.
Present status:
Routinely reported in national and international statistics.
Measurement method:
N/A.
Associated measurements:
Agricultural production, freshwater NO3.
Action required:
TBD.
FIRE AREA
Units of measure:
km2.
Users:
BGC, climate and eco-system modellers.
Rationale:
Burning of biomass is important because of the resultant release of a wide variety of trace gas species and aerosols. These affect regional and global atmospheric composition and in many regions such as the African savannas and boreal forests are crucial in affecting the vegetation succession.
Many fires are anthropogenic but lightning strikes are also an important factor. Fires are also an important indicator of land use and land cover change associated for example with deforestation.
Frequency of measurement:
Daily images are required that can be composited over time to give at least global monthly images.
Spatial resolution:
For the global distribution of fires, a 1 km resolution is probably sufficient, but their small areal extent means that reliable estimation of the areal extent of fires using the thermal bands is difficult. Information on the extent of fires can also be obtained through observations of the resultant fire scars. For global modelling purposes, these observations need to be reduced to much coarser cell sizes to be of manageable size.
Accuracy/precision required:
Accuracy: within 10% of the actual value;
Precision: within 5% of the actual value.
R and D needed:
Field campaigns are needed allowing surface and near-surface observations of fire processes linked to satellite observations.
Present status:
Information on the nature, distribution, timing and impacts of fires is currently poor. Using the AVHRR sensor, it appears likely that at least the broad global distribution and seasonal timing of fires can be determined. However, the coarse resolution of the sensor and the fact that the bands easily saturate in the presence of fire will limit the use of the resultant product. The MODIS sensor of EOS has the spectral bands to measure fire temperature much more accurately.
Measurement method:
Tiers 1-5: aircraft or satellite remote sensing.
Associated measurements:
Emissions of gases and particles; fire temperature.
Action required:
Current efforts through the IGBP to generate fire products should be supported as should research using new sensors with the capacity to monitor fire properties more reliably.
FIRN TEMPERATURE (ICE SHEETS, ICE CAPS, GLACIERS)
Units of measure:
Degrees C.
Users:
Modellers of sea level change, specialists for climate change detection.
Rationale:
Areas of cold firn predominate at high latitudes and under dry climatic conditions. In such areas, global warming primarily influences the distribution of temperature at depth. Most boreholes in Greenland, the Antarctica and on polar glaciers show strong deviations from steady state temperature profiles as a consequence of century warming. Increasing atmospheric temperatures can lead to additional melt water runoff to the sea.
Frequency of measurement:
Decadal.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
± 0.1 °C.
R and D needed:
The distribution of cold firn areas, outside the two large ice sheets should be modelled on the basis of glacier inventory information, topographic and climate data.
Present status:
Borehole temperature measurements in cold firn areas are commonly carried out in connection with ice core projects. Corresponding data need systematic collection and intercomparison. Borehole measurements at selected sites should be repeated at decadal time intervals.
Measurement method:
Tiers 2-3: temperature profiles in boreholes.
Associated measurements:
Determination of accumulation rates from ice cores recovered during drilling as described under the variable Ice Sheet Mass Balance.
Action required:
Organize systematic collection and interpretation of existing borehole data from cold firn areas.
GLACIERS AND ICE CAPS
Units of measure:
Change in ice thickness (m per year), glacierized area (km2 per year) and glacier length (m per year); energy flux (derived).
Users:
Change detection analysts, climate impact modellers, regional analysts and planners (hydrology, tourism, natural hazard).
Rationale:
Fluctuations of mountain glaciers are expressions of changes in the energy and mass balance at the Earth/atmosphere-interface. Glacier mass balance is the direct signal, whereas glacier length change is a strongly-enhanced, but also indirect, filtered and delayed signal of climate change. During the 20th century, glacier mass losses typically amounted to a few decimetres ice thickness per year. This corresponds to an additional energy flux towards the earth surface of a few watts per m2 and, thus, roughly corresponds to the assumed anthropogenic radiative forcing. Glacier shrinkage in some mountain areas such as the European Alps seems to have reached the "warm" limit of Holocene (pre-industrial) variability. Further evolution of such ice surface masses will constitute a key element of climate change detection and climate impact assessment.
Frequency of measurement:
Direct mass balance: annually;
Glacier length change: annually to once every 10 years;
Glacier inventories: every 30 to 50 years.
Spatial resolution:
Tiers 1, 2, 3, 4 and 5.
Accuracy/precision required:
Dependent on application and temporal resolution. High precision (0.1 to 1 m) for change detection, 100 m for model validation (inventories).
R and D needed:
Systematic application of remote-sensing products (air photos, high-resolution space imagery) should enable global coverage to be reached with respect to length/area change and inventory data. Analysis and worldwide intercomparison are needed to investigate the representativeness of the few existing long observational series. Special techniques such as laser altimetry combined with kinematic GPS or gravimetry have to be used in connection with measuring large glaciers and their effects on sea level.
Present status:
Standardized data are being collected and distributed by the World Glacier Monitoring Service (UNEP/GEMS, UNESCO, FAGS/ICSU, Haeberli, 1995). Mass balances are being measured on about 50 relatively small glaciers and glacier length changes are observed for about 1 000 glaciers. Some of these observational series reach back into the past century and can be reconstructed for the Upper Holocene. Detailed glacier inventories are available in numerous regions of the Americas, Eurasia, Africa and New Zealand. Concepts have recently been developed for combined (spatial and temporal) interpretation of inventory, length change and mass balance data. Large glaciers are important melt-water contributors to the ocean but are not well documented. Information from the Southern Hemisphere is still strongly limited.
Measurement method:
Tiers 1-3: direct measurements;
Tier 5: spectral radiance; microwave or laser altimetry.
Associated measurements:
Hydrological and meteorological data in glacierized catchments; sea level.
Action required:
Highest priority is on the continuation of long-term mass balance measurements and glacier length observations;
Better use should be made of remotely-sensed information on glacier size in areas of difficult access;
Baselines for long-term monitoring (surface elevation changes along representative flow lines of large glaciers in Alaska, Canada, Patagonia and the Himalayas) should be established using laser altimetry in combination with kinematic GPS;
Mass balance programmes should be strengthened in the Southern Hemisphere.
GROUND WATER STORAGE FLUXES
Units of measure:
Observation units = depth to water table;
Estimated flux units = net water volume change over observation period.
Users:
Hydrologic modellers; water resource planners; ecological modellers; climate change modellers.
Rationale:
The major portion of the world's fresh water occurs as ground water. Both recharge and discharge rates of ground water are affected by climate, albeit in complex and generally not instantaneous ways. Depletion of ground water stored in aquifers around the world (i.e., ground water mining) is due to withdrawals exceeding recharge. Climate change toward hotter and drier climates would encourage withdrawals for water supply and irrigation. As such, ground water depletion is a critical component for assessing impacts of climate change.
Recent research indicates that aggrading temperate forests and forest soils sequester significantly larger amounts of carbon than previously thought, and may be an important repository of the "missing carbon". The mobilization of such carbon through the hydrologic system, especially shallow aquifers, is poorly understood and may be a key mechanism controlling carbon exchange between terrestrial hydrological and ecological systems and the atmosphere. The documentation of groundwater levels and chemistry, in conjunction with surface water conditions, is necessary to quantify the role played by terrestrial hydrologic systems in carbon sequestration and mobilization.
Frequency of measurement:
For tiers 3 and 4, observation wells should be revisited at a minimum of once per year; For tiers 1 and 2, observations are needed monthly in conjunction with storm events.
Spatial resolution:
For the IOS, a representative sample of the world's aquifers needs to be developed;
For research purposes at tiers 1 and 2, it is dependent on the question being asked, but it is estimated the number will be between 10 and 100 per watershed.
Accuracy/precision required:
<1% of true depth.
R and D needed:
Improved understanding and modelling of ground water-surface water interactions, including water quality, fate and transport of chemical weathering products, and solute transport of radiatively active trace gases are required.
Present status:
Although global estimates have been made and published, there are major gaps in the data.
Measurement method:
Direct measurement at all tiers.
Associated measurements:
Precipitation, discharge.
Action required:
An inventory of existing programmes is needed so that a workshop of experts can select an IOS;
Ensure continued viability and integrity of long-term, process research, watersheds.
ICE SHEET GEOMETRY
Units of measure:
km and m.
Users:
Coupled ice-sheet climate modellers. Atmospheric modellers.
Rationale:
In the short term, changes in ice sheet mass balance may result from changes in surface balance, or changes in conditions at the margins, such as break-up of the ice shelves. The prediction of the response of the sheet to changes at the margins depends on the ice rheology, the ice temperature distribution, and the surface and basal conditions. The surface topography and velocities are boundary conditions on the modelling of ice sheet flow, and inferences as to the condition at the bed. The geography of the interior of ice sheets is still poorly known.
Frequency of measurement:
Decadal.
Spatial resolution:
100 m2.
Accuracy/precision required:
± 10 m.
R and D needed:
The performance of interferometric radar for determining topography and velocity, both absolutely and relatively, needs extensive investigation.
Present status:
At present, only coarse resolution (10 km) and poor accuracy (5 m) elevation models of ice sheets are available, provided by radar altimetry, and often these models are extremely poor at the ice sheet margins.
Complete radar imagery of Greenland is now available, and complete imagery of the Antarctica will become available with Radarsat (1995). Substantial proportions of the Antarctica have been imaged with ERS-1 (1991-1995).
Research in the capabilities of interferometric SAR commenced with ERS-1 and will continue with Radarsat. The eventual performance is yet to be determined.
Surface velocities derived from Landsat imagery have been provided for elements of the western Antarctic coast. These efforts remain labour-intensive.
The EOS laser altimeter (2004) will provide 100 m, 10 cm topography over much of the ice sheets, although the effect of atmospheric conditions on sampling is uncertain.
Measurement method:
Tiers 2-3: field surveys;
Tier 5: spectral radiance (high resolution) and altimetry.
Associated measurements:
The health of the Antarctic ice shelves, many of which are abruptly decaying in the Antarctic Peninsular, may be monitored with repeated imagery. In principle, the calving rates of icebergs can also be estimated, although the task is labour-intensive.
Action required:
It is important to stress that remote measurements cannot provide temperature profile data, and bed conditions may only be inferred. A continuation of present surface investigations, particularly of the western Antarctic shelf/stream system, is essential to providing better predictions of dynamic response of the ice sheet margins to atmospheric and oceanographic conditions.
ICE SHEET MASS BALANCE
Units of measure:
kg yr-1, kg m-2 yr-1.
Users:
Modellers for predicting climate change and in the estimation/prediction of global sea-level change.
Rationale:
The mass balances of the two large ice sheets of the Antarctic and Greenland is the difference between the mass gained by snow accumulation and that lost by ice ablation and calving. On short time scales, variations in surface balance as driven by atmosphere processes determine the mass balance of the ice sheets as a whole. Present ability to monitor changes in surface balance (snow accumulation and ice ablation) is limited by the paucity of observations. This uncertainty constitutes the largest uncertainty in determining the sources of the observed rise in eustatic sea level. In the event of global warming, the changes in the balances of the two sheets may largely cancel; Antarctica will experience increased precipitation; Greenland increased ablation. The uncertainty in this prediction is also large, perhaps as large as 0.5 mm yr-1 per degree C of warming. The mass balances of these two extensive ice sheets is also an important uncertainty in the interpretation of the observed rates of change of the Earth's rotation and gravity field.
Frequency of measurement:
Decadal.
Spatial resolution:
100 km2.
Accuracy/precision required:
5 × 103 kg yr-1.
R and D needed:
At decadal intervals, satellite radar altimetry can provide sufficient measurement accuracy to usefully constrain ice sheet mass balances over the interior, low gradient regions of ice sheets. At shorter time intervals, present accuracies are not sufficient, although this may change with improvements in orbit reconstructions.
Over the steeper gradients associated with the margins of ice sheets, ocean radar altimeters provide little useful data. Although limited in area, the ice sheet margins have the largest mass turnover, and are also those regions most affected by atmospheric and oceanographic changes. Different technical solutions are required for these areas. While a laser solution is currently in development, alternative radar solutions should be pursued. It is not clear at present that atmospheric conditions will permit adequate sampling of the margins.
The remote-sensing of accumulation remains a critical research goal.
Present status:
Radar altimetry observations of Greenland and the Antarctica to 72 degrees of latitude were made by Sea Satellite (Seasat) (1978) and Geodesy Satellite (Geosat) (1985-1989). Radar altimetry coverage to 82 degrees of latitude have been made by ERS-1 and will continue with ERS-2. To date, elevation change measurements are possible with these data to an accuracy of 0.5 m, principally limited by orbit reconstruction error. The Seasat/ERS data set should usefully constrain the balance of Greenland. At present, it appears unlikely these data have a time interval long enough to usefully constrain the Antarctic balance. The European Space Agency's Environmental Satellite (ENVISAT) radar altimeter (1998-2003) will repeat radar altimeter coverage to 82 degrees of latitude. The EOS polar altimeter will provide laser altimetry of ice sheets to 86 degrees of latitude.
In regions of net surface accumulation, the surface balance may be determined from ice cores. Accumulation is relatively well constrained in the margins of western Antarctica and parts of eastern Antarctica. In Greenland, accumulation is relatively well known in the south. It is probable that a considerable amount of accumulation data exists in institutions around the world. The lack of surface melting in the Antarctica makes determining the surface balance simply a question of determining accumulation. In Greenland, surface ablation accounts for half the mass loss from the sheet as a whole. However, few measurements exist of surface ablation, and these are located at the coast.
While surface balance cannot be determined from satellite observations, passive and active microwave sensors can provide considerable information on the state of the surface layers. Regions of permanently dry snow, surface melting and bare ice regions can be distinguished. Variations in the equilibrium line that separates regions of net accumulation from regions of net ablation can be determined by satellite observations.
Measurement method:
Tiers 2-3: field surveys, GPS;
Tier 5: altimetry.
Associated measurements:
Accurate change measurements with altimeters are possible only through the continued operation of the global satellite tracking networks. The lifetime of satellites (3-5 yrs.) will allow sampling of the interannual variability of elevation.
Action required:
The continued operation of radar/laser altimeter satellites, with high inclinations (>80 degrees) and repeat cycles in excess of 30 days, is essential to the long-term monitoring of ice sheet mass-balance;
Research in radar solutions to the monitoring of ice sheet margins should continue. It is not clear at present how the planned laser altimeter will perform at the ice sheet margins;
Shallow cores (200 yr) over the interior of the Antarctic and Greenland Ice Sheets should be obtained in a systematic fashion. The relatively uniform climate of eastern Antarctica makes even sparse sampling useful. In Greenland, a higher spatial density (104 km2) is required;
The surface measurement of surface ablation, particularly in Greenland, should be extended, both in altitude and geographical distribution;
The calculation of water flux divergence over the ice sheets.
LAKE AND RIVER FREEZE-UP/BREAK-UP (Timing)
Units of measure:
Julian date.
Users:
Assessors of climate change detection and impact.
Rationale:
Conventional ground-based observations of the dates of lake and river freeze-up/break-up are well correlated with air temperature during the transition seasons. As a change indicator, freeze/break-up dates change by approximately 4-7 days for every degree (Celsius) change in air temperature. Changes in lake and river ice cover will have not only ecological effects on freshwater systems but can also have significant economic effects, for example, by affecting ice road transportation.
Polar lakes with permanent ice cover may be particularly sensitive indicators of high-latitude change.
Frequency of measurement:
Daily observation in spring and fall.
Spatial resolution:
Two hundred medium-sized lakes (25 to 100 km2) and selected large lakes geographically distributed across middle and high latitudes.
Accuracy/precision required:
Date +/- 1-2 days. Date of complete lake freeze-up/break-up is required.
R and D needed:
Satellite mapping capability has been demonstrated but not yet implemented for Arctic lakes. SAR can be used to identify lakes which freeze to the bottom; passive microwave data can be used to map ice cover/open water of large lakes.
Present status:
Ground observations for many lakes and rivers in North America, Russia and European countries; inadequate metadata and archives. Use of visible band satellite imagery is limited by cloud cover.
Measurement method:
Tiers 3-4: in site observations;
Tier 5: spectral radiance, microwave backscatter.
Associated measurements:
Water temperature and air temperature, wind velocity, precipitation, and snowfall at adjacent climate stations.
Action required:
A central archive or several regional archives (e.g., North America, Northern Europe, the Russian Federation, South America, Himalayas) are needed;
A comparison of conventional methods with satellite-derived time series is needed;
A compilation of existing records is needed so that a long-term set of "GCOS" lakes can be selected.
LAND COVER
Units of measure:
Land cover class.
Users:
International, regional, and national agencies responsible for agricultural, forestry, and urban/industrial planning. GCM modellers, BGC modellers (global, regional), ecosystem modellers, climate impact modellers; also regional analysts/planners.
Rationale:
Changes in land cover drive changes in the climatic system and are affected by climate change either directly or through human response. Land cover determines water, energy, and trace gas balances and changes in cover have significant effects on carbon flux. (For example, estimates are that tropical deforestation contributed between 2.0 and 2.8 Gt C per year during the late 1980s, contributing 25-35% of total biospheric emissions of carbon.) Climate change could significantly expand the area of land suitable for cultivation in northern latitudes, simply through an increase in the length of the growing season. Increasing the length of the growing season could significantly extend the northern limits of cultivation, but drought may restrict crop production and forest boundaries at lower latitudes. Comprehensive, standardised characterization of land cover is needed to drive models that simulate regional changes in cover and to understand the feedback to climate.
Frequency of measurement:
Global mapping: every 5 years;
Annual observations are needed in indicator/sensitive areas to determine short-term fluctuations.
Spatial resolution:
IOS: 1 km for global coverage; 30 m for regional to local coverage; Post-IOS: 0.25-0.5 km for global, 30 m for regional to global.
Accuracy/precision required:
Global accuracy of 80% may be sufficient for the operation of GCMs, but for other purposes, such as estimating changes in carbon stocks and detecting cover boundaries, accuracies of better than 95% are required.
R and D needed:
Continued effort is required to consolidate global cover mapping using existing methods. More detailed research is required to improve discrimination of more subtle differences in cover and for small areas of heterogeneity. Wetlands are a special case which would benefit from improvements in remote-sensing techniques. Improvement is needed in detecting and monitoring fires.
Present status:
Global coverage can be obtained from the Global 1 km AVHRR data set (10-day composite images over 18 months) which was completed in 1995. A methodology for regional land cover mapping was demonstrated in North America.
High resolution coverage is provided by high resolution satellite data (Landsat, SPOT, etc.). The methodologies for mapping have been well demonstrated. Methodologies for change detection continue to be developed, with operational techniques expected in 2-3 years. Significant regional coverage is feasible with newly-obtained and archived data. A good representation of wetlands distribution is especially important for BGC studies. At the global level, it will be more reliably derived by combining satellite data outputs with soil data.
Measurement method:
Tiers 1-3: field mapping; spectral radiance (high resolution);
Tier 4: field observation;
Tier 5: spectral radiance (high to medium resolution).
Associated measurements:
Land use.
Action required:
Several land cover maps are available. However, there are large disagreements between them due to differences in methodologies used to prepare them. What is needed is an adequate characterization of the spatial distribution of actual vegetation cover. For historical periods, current methods probably cannot be improved. However, for the recent past, present, and future, land cover will be best provided by remote-sensing. For spatial extrapolation purposes, the spatial resolution, in general does not have to be great. 1-5 km is finer than that which most investigators currently use but not finer than the scale over which significant variations in vegetation processes occur. It is important to distinguish land cover categories along criteria related to carbon storage and turnover. Olson, et al. (1983) suggested a scheme that resulted in 30 or so categories that included possibilities for mixed pixels at an 0.5 degree resolution. At a finer spatial scale and using different methods or tools, other classification schemes will undoubtedly be necessary. However, it is thought that 20-40 categories of land cover would be adequate. For wetland categories that have special carbon dynamics, higher spatial resolution may be required. A rationale for change detection and parameters is required.
Endorse completion of the initial global 1 km AVHRR data set;
Promote continuation of the Global 1 km AVHRR project to provide a continuing series of global observations;
Promote validation of cover mapping methodology for 1 km data in various regions;
Encourage development of robust change detection algorithms;
Promote expanded use of high resolution data for land cover mapping for large areas (national to regional levels);
Promote generation of global data sets with higher spatial resolution from existing or new sensors.
LAND USE
Units of measure:
Land use class.
Users:
International, regional, and national agencies responsible for agricultural, forestry, and urban/industrial planning. BGC modellers, ecosystem modellers.
Rationale:
One of the most controversial components concerning the global carbon cycle concerns the magnitude of release and uptake of CO2 resulting from man's activities. Currently, the most authoritative accounting of these fluxes is based on interpretations of land use statistics compiled by FAO. A more objective system that can be updated annually may be possible using remote-sensing of land cover change (perhaps combined with agriculture and forestry production statistics, socio-economic spatial data such as population and transportation). Landsat is one of the most appropriate data streams for land use monitoring. While it is still a research topic on how to use this data to produce a land use product, it is important that this data stream continues to be archived and kept accessible. The type of land use which can be sustained under future climatic conditions is also one of the most troubling questions of climatic change from the point of view of human impact and economic terms, and from its impact on the climate itself.
Frequency of measurement:
Once every 5 years.
Spatial resolution:
5 m - 1 km depending on the spatial heterogeneity of land use; at least 30 m for many regions.
Accuracy/precision required:
· TBD.
R and D needed:
Development of regionally-specific relationships between land cover and land use;
Definition of the lowest acceptable spatial resolution (also regionally specific, depending on land use heterogeneity);
Development of reliable procedures of inferring land use from land cover, on a regional basis.
Present status:
Land use maps have been completed for many specific areas of the world. These efforts were based on various data sources and are not generally compatible with each other. A consistent, hierarchical classification system is presently under development; this is to be encouraged.
Measurement method:
Tiers 1-4: field surveys, supported with remote-sensing images;
Tier 5: inferred from spectral radiance (high resolution) (experimental).
Associated measurements:
Standing surface water, ground water, land cover.
Action required:
Complete the development of land use classification scheme;
Identify areas of the world for which acceptable and accessible products exist; compile these into a consistent database;
Carry out pilot studies for filling the critical gaps in coverage using a land-cover mapping procedure and the relationships between land cover and land use;
Validate preliminary products.
LEAF AREA INDEX (LAI)
Units of measure:
m2/m2.
Users:
Climate modellers, weather forecast modellers, ecosystem modellers, BGC modellers, modellers of atmosphere-ecosystem interactions.
Rationale:
The most suitable definition of LAI is half the total projected green leaf area (one-sided area for broad leaves) in the plant canopy per unit ground area (Chen and Black, 1992). LAI describes a fundamental property of the plant canopy in its interaction with the atmosphere, especially concerning radiation, energy, momentum and gas exchange (Monteith and Unsworth, 1990). Leaf area plays a key role in the absorption of radiation, in the deposition of photosynthates during the diurnal and seasonal cycles, and in the pathways and rates of biogeochemical cycling within the canopy-soil system (Bonan, 1995; Van Cleve, et al, 1983). Various soil-vegetation-atmosphere models and BGC models use LAI (Sellers, et al., 1986; and Bonan, 1993a). Globally, it varies from less than 1 to above 10 but also exhibits significant variation within biomes at regional, landscape, and local levels.
Frequency of measurement:
IOS: monthly;
Post-IOS: product to be issued every 10-30 days.
Spatial resolution:
IOS: 1-8 km for global applications;
Post-IOS: 0.25-1.0 km.
Accuracy/precision required:
IOS: 15-25% of true value;
Post-IOS: 5-15% of true value.
R and D needed:
Development of methods in various biomes and validation. Improvement of methods to increase robustness and to reduce dependence of ground data. Improvements in the canopy radiation models used as the basis for inferring FPAR from satellite measurements and for relating satellite values to various levels within the canopy.
Present status:
Similar to FPAR, a global LAI data set has been produced (Sellers, et al., 1994). LAI in the data set is calculated from FPAR which is treated as a linear function of the simple ratio. Different relationships between LAI and FPAR were used for two broad vegetation types. The validation of the data set has not yet been done (Hall, et al., 1995). There have been many published results on the relationship between vegetation indices and LAI for agricultural crops and grasslands, but similar results are limited for natural ecosystems.
Measurement method:
Tiers 1-3: destructive sampling; allometric equations; indirect (light penetration-based estimates);
Tier 5: spectral radiance (high or medium resolution).
Associated measurements:
LAI estimation from satellite data requires ground data for validation and testing for bias. Satellite data must be corrected for atmospheric effects, thus requiring additional information on the state of the atmosphere (especially water vapour, aerosols, ozone). Non-destructive (optical) measurements are the preferred approach for obtaining ground measurements (Chen and Cihlar, 1995b).
Action required:
Complete testing of non-destructive ground methods in various biomes, especially needleleaf forests;
Continue development of prototype LAI products at various spatial resolutions. Test ability of satellite data to produce LAI products in different parts of the phenological cycle. Test use of products in models;
Produce time series of LAI products at various spatial resolutions and coverage, from regional to global. Transfer methodologies to new data sources and optimize;
Continue testing and use of products in models.
METHANE FLUX (CH4)
Units of measure:
ppb.
Users:
Climate change and ecosystem modellers and researchers. Science policy advisors nationally and globally.
Rationale:
Methane is one of the most important radiatively active atmospheric trace gases with the potential to affect climate significantly within the next century. Atmospheric concentrations have risen rapidly since records began in 1978 (0.8-2.0% per year), but there are signs of decline in the 1990s. Trends are well documented but the magnitude of sources and sinks, and their individual trends, are less well known. Terrestrial sources are mainly ruminants, rice paddies, northern wetlands/peatlands and landfill. Changes in temperature and precipitation can have significant effects (both positive and negative) on CH4 emissions through changes in permafrost and waterlogging. Decomposition of methane hydrates in continental shelf regions and landfill emissions are important additional sources needing quantification and monitoring. Linked observations at different scales are essential to detect trends, refine models and assess policies.
Frequency of measurement:
Continuous or monthly observations of atmospheric concentration. Cumulative or weekly/monthly 24 hr samples at tier 3; Continuous measures over short-term seasonal phases at tiers 1 and 2.
Spatial resolution:
Tiers 1 and 2;
Metre plots or km transects for tier 3 representative sites;
Intensive profile measurements over km fetch on tiers 1 and 2;
Accuracy/precision required: 95% accuracy, but quantification of spatial variation is more important than analytical error.
Accuracy/precision required:
± 5%.
R and D needed:
Among the many requirements in this critical subject are: development of improved satellite and airborne sensors for concentration and eventually for flux determination; development of long-path detection systems for ground-based, landscape scale flux measurements; extensive field measures (with experimental treatments) to assess, then monitor, emission and absorbtion rates. Improvement in satellite definition of wetlands/peatlands and detection of change in temperature/moisture/permafrost regimes are critical.
Present status:
Atmospheric concentration monitoring is reasonably well established through the NOAA/CMDL air sampling network. Field-based observations are a completely sporadic and poorly-distributed research activity despite the concentration of effort in wetlands/peatlands and tundra. Rice paddy observations are improving considerably through the leadership of the International Rice Research Institute (IRRI). Estimation of sink/source relationships and of change have very high errors due mainly to inadequate sampling.
Measurement method:
Tiers 1-3: eddy correlation; chambers; aircraft-based eddy accumulation;
Tier 5: spectral radiance (future; to be demonstrated).
Associated measurements:
Quantitative definition of land cover and its derivative, land use, are the important basis for extrapolation from spot to large area. Remote-sensing of wetlands/peatlands is critical but difficult. Systematic associated observation of climate, moisture/redox, vegetation and soil conditions (especially heterogeneity in tundra) are essential for calculation, comparisons and extrapolation.
Action required:
Support for IRRI to establish rice paddy monitoring;
Establishment of wetland/peatland monitoring system for CH4 flux and change in conditions;
Encourage development of techniques and systematic recording of CH4 flux over wetland/peatland landscapes.
NECROMASS
Units of measure:
kg/m2.
Users:
Ecologists modelling transient responses of forests to rapid climate change; ecologists and climate modellers estimating terrestrial carbon stocks and carbon fluxes; decision-makers concerned with international measures to mitigate atmospheric carbon increases.
Rationale:
Necromass consists mainly of plant litter. It is usually on the soil surface or in the soil, but some may take the form of standing or attached dead material. Much of the transient or lag in response to rapid climate change by forest ecosystems can be estimated by the difference between tree natality and tree mortality. Annual necromass increments result from individual tree mortality within stands and from larger-scale disturbance and die-back events (fires, insect infestations, disease infestations, wind throw). In addition, a significant portion of the carbon stocks which comprise terrestrial carbon stored in forest and non-forest communities can be found as necromass which must be documented if the terrestrial carbon cycle is to be understood and predicted, and if carbon status in geopolitical regions is to be documented.
Frequency of measurement:
Annually.
Spatial resolution:
Tiers 1, 2 and 3.
Accuracy/precision required:
± 10%.
R and D needed:
No known means of remote detection; early attempts using European forest die back failed because of forest management aimed at sanitation removals; die-back detection in north-eastern U.S. required very fine-scale spatial measures in 100% mortality. In the longer term, development of remote-sensing techniques will be necessary.
Present status:
Poorly known and very little documentation of spatial and temporal distributions.
Measurement method:
Tiers 1-4: field observations.
Associated measurements:
Above-ground and below-ground biomass, vegetation structure, vegetation roughness.
Action required:
Systematic measurement at the scale of points, stations and experiments;
Eventual development and employment of remote-sensing techniques to document necromass distributions.
NET ECOSYSTEM PRODUCTIVITY (NEP)
Units of measure:
moles C/m2; kg C/ha/yr.
Users:
Vegetation ecologists, ecosystem modellers, global vegetation modellers, decision-makers.
Rationale:
Like NPP, NEP includes the annual sum of carbon fixed by daytime photosynthesis and of carbon released by night time respiration of fixed carbon. Unlike NPP, NEP also includes carbon released by maintenance respiration inside plants, by microbial respiration of organic compounds in dead plants and branches, in surface leaf litter, and in soils, and respiration/loss of grazed living plant material by other respiring animals. Thus, NEP is a measure of the residual annual carbon incremented to the long-term carbon stocks of any given region or location. As such, it is a critical variable in determining long-term behaviour of the terrestrial carbon cycle, and in defining changes in long-term carbon stocks.
Frequency of measurement:
The variable itself needs to be measured annually; however, the means by which NEP is calculated (e.g., with eddy-flux measurements) may dictate continuous or hourly measurements, or (with cuvette methods that separate canopy and soil fluxes) monthly estimates of carbon flux.
Spatial resolution:
Tiers 1, 2 and 3.
Accuracy/precision required:
± 25% instantaneous point measurements, ± 10% for annual and landscape-level values.
R and D needed:
Early methods involving repeated annual sampling of annual tree increments (as derived from tree rings and root samples collected every metre of length or of whole tree harvests), of leaf and seed fall, and of soil respiration clearly were too inaccurate and labour-intensive for global applications. The eddy-correlation technique of carbon flux measurement is only now being evaluated with intensive field studies; this work needs to be pushed to completion while other, perhaps more direct, methods are developed. For the latter, three main areas need exploration: (a) instrument system development; (b) spatial representativeness and integration; and (c) internal constraints and redundancies between C and other measurable variables. Ideas for (b) include repetitive airborne transects, radon, and 13C; for (C), C:N and C:H2O relationships, and annual comparison with biomass change.
Present status:
At present, eddy correlation measurements are being undertaken in few locations worldwide and their validity in measuring carbon flux is uncertain at best. This is of no significance in terms of the minimum needs for carbon flux measurements and resulting NEP estimates.
Measurement method:
Tiers 1-3: eddy correlation; field measurements (computed at tier 5).
Associated measurements:
Above-ground and below-ground biomass, necromass, vegetation structure, vegetation roughness; in addition, photosynthesis, leaf respiration, NPP and leaf area estimators (LAI, FPAR, etc.) are components of NEP.
Action required:
In the short term, evaluation of the eddy-correlation instruments and the overall technique of carbon flux measurement must be completed with intensive field studies aimed especially at defining spatial representativeness; if this work is successful in certifying the strengths and weaknesses of the method, data collection sites must be developed. At first, these may only be at the level of land experiments and centres, but eventually they should be added at the level of stations and points;
For the longer term, more direct methods of measuring NEP should be developed.
NET PRIMARY PRODUCTIVITY (NPP)
Units of measure:
kg/m2.
Users:
Ecosystem modellers, BGC modellers, resource managers, planners.
Rationale:
NPP refers to the net carbon uptake by the ecosystem in one year. It is defined as the difference between total carbon uptake through photosynthesis and losses (through maintenance or growth respiration). NPP is a sum of two components, above (ANPP) and below (BNPP) ground. The distinction between ANPP and BNPP is made because of the differences in the type and rates of processes involved and in our ability to measure these two parameters.
NPP is a fundamental property describing ecosystem performance. It depends on the characteristics and interactions of the vegetation, soil, and weather, and integrates these effects in a spatially explicit manner. It provides a record (e.g., in annual growth increments of woody species) of these interactions. Because of its direct relationship to atmospheric CO2 it plays a strong role in the global climate system. Seasonally it accounts for the latitudinal dependence of the atmospheric CO2 concentration which can be detected from satellites (Fung, et al., 1987). On the annual basis, it accounts for the land portion of the carbon sink and may explain the missing carbon (Tans, et al, 1990). In the past, NPP has not been measured systematically or frequently because of the limitations of measurement methods. Thus, only typical values are available for various ecosystems, and the interannual variability is not known except for a limited number of specific sites. Recent advances in BGC modelling (e.g., Bonan, 1993b) and satellite data applications (e.g., Prince, 1991(b); Prince, et al., 1994; Goward, et al., 1985; Goward and Dye, 1987) offer the prospects for global mapping of NPP.
Frequency of measurement:
Ground measurements: once a year if biomass sampling approach is used; daily if eddy flux correlation is used;
Satellite: daily estimates of photosynthesis and respiration to obtain annual total.
Spatial resolution:
Ground measurements: Tiers 1, 2 and 3;
Satellite measurements: IOS: 1-8 km. Post-IOS: 0.25-1 km.
Accuracy/precision required:
± 10%.
R and D needed:
Validation of models in various biomes and years, at various spatial resolutions from stand/patch to landscape. Testing models with various input parameters and degrees of process representation. Production and validation of NPP estimates from satellite data, including estimation of ANPP and BNPP components.
Present status:
Models with various degrees of complexity have been developed and tested, principally in forest and grassland ecosystems (Bonan, 1991, 1993b). The use of satellite data in conjunction with models has been demonstrated in specific situations (Running and Nemani, 1988; Royer, et al., 1995, Runyon, et al., 1993) but no rigorous independent tests have been carried out at continental or larger scales. Satellite data sets are being developed through the EOS Pathfinder programme (Eidenshink and Faudeen, 1994; James and Kalluri, 1994) which should be used to carry out rigorous tests of NPP estimation using satellite data. Depending on the model, ANPP or NPP will be obtained in this manner. The BNPP component may be obtained through the detailed process models used or by establishing relationships between ANPP and BNPP using environmental parameters available at continuous fields.
Measurement method:
Tiers 1-3: eddy correlation; field measurements; chambers (modelled at tier 5).
Associated measurements:
Ground measurements on NPP are required to validate the results from modelling. Meteorological data, soil data and plant parameters are required as input variables for process-based models.
Satellite measurements: depending on the model needs, FPAR, Incident Photosynthetically Active Radiation (IPAR), Absorbed Photosynthetically Active Radiation (APAR), and LAI products are required.
Action required:
Continue producing data sets and improved models for testing accuracies of NPP estimation at regional to global scales;
Carry out rigorous tests of various approaches in different biomes;
Initiate pilot programmes for NPP estimation at study areas representing various biomes of the world. Complement these with samples of ground methods for assessing accuracy and checking for bias or drift;
Continue preparing data sets for global application and initiate retrospective processing of multi-year series (Pathfinder AVHRR) for global NPP estimation.
PERMAFROST - ACTIVE LAYER
Units of measure:
Depth to permafrost table (cm) if possible combined with soil temperature (°C) at 15 cm depth.
Users:
Change detection analysts, climate impact modellers, regional analysts/planners (hydrology, engineering), plant ecologists in cold areas (polar and high mountains).
Rationale:
Permafrost areas at high latitudes and high altitudes would be among the regions most heavily affected by continued or accelerated warming. Active layer changes are the direct/undelayed response of permafrost to changing surface conditions. Thaw destabilization of perennially frozen ground, therefore, has strong effects on hydrology, vegetation, slope stability, surface albedo etc., but the processes involved are complex and not well understood. Moreover, thawing of organic layers could release additional amounts of methane.
Frequency of measurement:
100 points at weekly intervals after snow melt at occupied sites, if possible combined with recording at 15 cm depth of active layer temperature. Statistically representative measurements/soundings in late summer (maximum thaw depth) or installation of thaw tubes, at non-occupied sites.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
0.01 to 0.1 m for a statistically significant number of regularly arranged points following recommended standards of the IPA (Circumarctic Active Layer Monitoring (CALM) protocol), or 0.01 m with permanent thaw tubes.
R and D needed:
Surface energy balance, especially with respect to snow cover effects and vertical/horizontal heat and mass flux between the surface and the permafrost table, must be better understood.
Present status:
Despite its importance to a wide variety of physical and biological processes related to climate change, the development of the active layer has rarely been documented in a systematic, standardised fashion over large areas. The majority of the historical record on thaw depth has been obtained by physically probing for the permafrost table - a low-cost method of nondestructive and areally extensive data collection in peaty soils and shallow active layers. Geophysical prospecting and/or shallow drilling with temperature monitoring or frost tubes must be used in coarse material and deeper active layers. A protocol for CALM has been prepared by the IPA and has already been implemented at 20 sites. To take advantage of complementary measurements, the CALM protocol should be initiated at most of the 25 International Tundra Experiment (ITEX) stations and some Arctic long-term ecological research (LTER) and TEMS sites. Where soil conditions permit, frost tubes should be installed. To assure circumpolar distribution some 10 additional new sites should be established; it is expected that existing permafrost sites in Canada and the Russian Federation can be enhanced for this purpose. Several new sites are required in the southern hemisphere including Antarctica.
Measurement method:
Tiers 1-4: field measurements (supported by remote-sensing images).
Associated measurements:
Permafrost thermal state (temperature measurement at 15 cm), snow cover area and snow water equivalent, vegetation structure, land cover, soil moisture, soil bulk density, ground-water storage fluxes, precipitation, radiation and topography.
Action required:
Institute protocols and recommendations of the IPA at key existing sites and at selected international programme sites (e.g., ITEX);
Select new sites where geographic gaps exist. It is expected that many of the gaps will be in high-altitude areas.
PERMAFROST - THERMAL STATE
Units of measure:
Temperature (°C) at depth (m), heat flow/energy flux (W/m2, derived).
Users:
Change detection analysts, climate impact modellers, regional analysts/planners (engineering, oil and gas prospection, natural hazards) in polar and high mountain areas.
Rationale:
Thermal conditions within permafrost at high latitudes and high altitudes are closely linked to changes in atmospheric and surface conditions. Corresponding changes in temperature distribution and vertical heat flow at depth take place over time scales of decades to centuries, and adjustment of permafrost thickness can take decades, centuries or even millennia depending on original permafrost thickness and ice content. Due to the important memory and filter functions as well as the long time scales involved with heat conduction in virtually impermeable ground, recent warming of near-surface temperatures in permafrost constitute a key signal of climate change impact with foreseeable long-term effects. Pronounced disturbances of temperature profiles related to 20th century warming can be observed in the uppermost 200 m of sub-arctic boreholes.
Frequency of measurement:
Weekly to monthly in the uppermost 20 m below the permafrost table, annually to once every 10 years for depths greater than 20 m.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
0.1C for near-surface temperatures and 0.05 to 0.1C for temperature profiles in deep boreholes.
R and D needed:
Borehole installations must be designed for long-term monitoring of temperature and vertical/horizontal deformation (creep, thaw settlement/frost heave). Thermal characteristics of the investigated ground material must be carefully determined using a combination of laboratory and in situ measurements (thermal adjustment after drilling, decay/lag of seasonal amplitudes with depth). Special emphasis should be on effects of unfrozen water near melting/freezing temperature.
Present status:
A number of sites are presently being monitored in North America (Canada, Alaska), in the Russian Federation and in some high-altitude areas (Tibet Plateau, Tien Shan Mountains, European Alps). Efforts are being undertaken by the IPA to systematically rescue, collect and distribute such data. Within the same framework, a circumpolar/northern hemisphere permafrost map at a scale of 1:10 000 000 was prepared and could be used for GCM-validation in polar regions.
Measurement method:
Tiers 1-4: borehole temperature measurements (continuing or periodical).
Associated measurements:
Snow cover, meteorological data, permafrost active layer, vegetation/surface changes.
Action required:
Develop criteria for selection of key existing sites to recommend continuance;
Establish new sites where geographic gaps exist, including high-altitude areas;
Temperatures at 15 cm and in boreholes greater than 20 m depth should both be implemented.
PLANT TISSUE N AND P CONTENT
Units of measure:
mg N/kg, mg P/kg for photosynthetic and structural tissues.
Users:
BGC modellers.
Rationale:
The plant biomass constitutes a major pool for C, N and P. To calculate these pools, and their changes over time, it is necessary to know the biomass and the elemental content of the biomass. The fraction of C content hardly varies and therefore need not be measured. The content of N and P is variable, and sensitive to the availability of these elements. Furthermore, the content of N has a major impact on the potential photosynthetic rate, respiration rate and digestibility of tissues. The N content of particular groups of plants is especially important in calculating ecosystem N inputs from biological nitrogen fixation. The 15N content of the leaf N, where available, indicates what fraction is derived from biological N fixation.
Frequency of measurement:
Twice per year, at the time of first full leaf expansion and at leaf fall.
Spatial resolution:
One sample for each species contributing at least 10% to the primary productivity of the site; to a maximum of 10 species in very diverse vegetation, one sample from a representative biological nitrogen fixing species where present.
Accuracy/precision required:
10%.
R and D needed:
None.
Present status:
N content is routinely analysed at many sites, but not globally observed. P content is less widely analysed.
Measurement method:
Tiers 1-3: direct measurements.
Associated measurements:
Plant biomass.
Action required:
TBD.
PRECIPITATION - ACCUMULATED (SOLID AND LIQUID)
Units of measure:
mm per time interval.
Users:
Widespread use by scientists, resource managers, modellers, planners, climate change detection and impact analysts, climate system analysts, and many others. Critical variable for assessing regional and global water supply and its variability with time. Characterizes the input of water into the entire hydrological system. Critical for a range of models including climate, weather, ecosystem, hydrological, biogeochemical models.
Rationale:
Precipitation includes fallen (liquid, solid, or rain, snow, hail) as well as deposed (dew, hoar-frost, rime etc.) forms. The observed variable is "Precipitation depth", which is defined as the depth of liquid water accumulated during a defined time interval on a horizontal surface. "Precipitation rate" is used for production of liquid and solid water by condensation and sublimation in the atmosphere during a time unit. Precipitation must be measured to determine the annual accumulation required for regional and global water balance estimation. Precipitation drives the land surface hydrology in the same way that incoming solar radiation drives the surface thermal regimes. It is therefore the key variable in the terrestrial hydrological cycle, surface water budget, and is essential for all vegetation growth. Reliable, high-resolution records of precipitation are critical as input for understanding and monitoring regional effects of climate variability and change, as input for hydrological models, and for validating global and regional climate models. Precipitation is also the primary determining factor for national and global resources and for crop production (forestry and agriculture). There are different requirements for precipitation data, depending on the use. For large-area applications ('macro-scale'), data are required with a coarser spatial and temporal resolution than for specific studies over smaller regions and/or where higher resolution is called for ('meso-scale'). For most of the applications (model verification, water cycle and budget), area averaged information on precipitation is required.
Frequency of measurement:
Macro-scale: daily (but note that the measurements are not time-coincident around the globe);
Meso-scale: hourly where available.
Spatial resolution:
For meso-scale analyses: 1 km up to 10 km;
For macro-scale analyses: 50 km up to 100 km.
Accuracy/precision required:
The raw data should be available, corrected for systematic errors. For both macro- and meso-scale studies: 0.1 mm should be the definition of precipitation as opposed to no precipitation.
|
Instrument precision: |
liquid precipitation: 0.2 mm or less; solid precipitation (snow): 0.6 mm or less. |
|
Accuracy: |
liquid precipitation: ± 1 mm for <= 20 mm; ± 5% for > 20 mm; snow: 5% of true value; (adjustment for systematic errors required). |
R and D needed:
Ongoing WMO work, with regard to the correction and standardisation of solid precipitation measurements, is endorsed. Assimilation of in situ measurements with remotely-sensed (radar and satellite) measurements is required. Adjusted precipitation estimates for liquid and solid precipitation need to be produced and archived. More reliable ways are needed of remote-sensing solid precipitation particularly in high latitudes. (Present efforts to derive precipitation rates focus largely on the tropics.)
Present status:
Work is ongoing in the GPCC to derive monthly precipitation fields; this work is endorsed for model validation purposes. Therefore, the incorporation of model-generated data in these fields is to be discouraged. GPCC can use approximately 6 000 stations of which data are disseminated worldwide via the GTS. This is only a subset of the total precipitation network. About 200 000 raingauge stations are operated in national networks. Studies on the accuracy of solid precipitation measurements are being finalized. Data of at least 40 000 stations are required for reliable global analyses on a monthly basis. Some countries are installing automated weather stations without sufficient solid precipitation measurement capability.
Measurement method:
Tiers 1, 2, 3: direct in situ measuring using precipitation gauges.
Tier 5: indirect remote-sensing techniques: Based on radiation theories, radar and satellite observations can be used to estimate area-mean precipitation. Remote-sensed precipitation data need to be adjusted to in situ observations.
Associated measurements:
Detailed topographic database to allow aggregation/disaggregation;
Metadata about reliability: altitude, instrument type; station description, climatological normals; meteorological conditions during the precipitation event: air temperature, dewpoint, precipitation type current weather, snow depth, wind speed. Information on quality control procedures and results, and on any applied corrections.
Action required:
There should be a concerted effort to use the data from existing sites. A very large number of these measurements exist, and efforts are probably better directed to obtaining existing data than to establishing new stations. Sites for inclusion in a data set should be chosen taking into account the location of discharge stations so that the two data sets can be used for river basin studies (in the tier framework, as applicable). Data from non-synoptic stations run by meteorological and hydrological services should be incorporated into the global precipitation database. They are often at more representative locations and are more numerous than synoptic stations.
On a more general basis:
Macro-scale studies: the first need is for 40 000 stations based on an objective analysis of the distribution of stations. We can use existing knowledge of actual versus desired stations based on information available to GPCC together with WMO criteria. These should give GPCC a basis for establishing guidelines for locating desired sites;
Meso/micro-scale studies: the network will depend on the purpose for which the data are to be used and the local climatic/topographic conditions. In most cases the request will be for all possible stations in the area concerned, even if this gives non-uniformity in the density of stations;
Pursue technological advances in the direct measurement of precipitation over oceans;
Continue using informal personal contacts to achieve the most comprehensive coverage possible until data delivery (continuation of delivery of data from countries supplying at the present, encouragement of countries not supplying data to do so), until routine data delivery to GPCC becomes more established;
Send a letter to all countries requesting that an update of the precipitation measurements be forwarded to GPCC, including the most recent data and all synoptic station data;
Prepare a statement to governments or Heads of States to give priority to regular and more comprehensive precipitation and runoff data submission to GPCC and GRDC, respectively;
Design a future data flow structure for precipitation data and products (end to end);
Autostations should measure both liquid and solid precipitation;
Precipitation measurements should be routinely corrected for systematic measurement errors, but the raw data should be available, too.
RADIATION - ALBEDO
Units of measure:
No unit.
Users:
Modellers and analysts concerned with climate, weather, ecosystem, hydrological issues.
Rationale:
As the ratio of the downwelling to upwelling shortwave radiative fluxes at the surface, surface albedo is fundamental to surface energy balance. Surface albedo and clouds are the two most important parameters modulating the Earth's climate, especially over land. Although it does not change temporally as much as clouds do, surface albedo undergoes considerable variation over space, and it shows strong seasonal cycles and inter-annual variability. Surface albedo data have been obtained from in situ observations (Ohmura and Gilgen, 1991) and from space-borne measurements by means of remote-sensing techniques (Li and Garand, 1994). The former are site-specific as they usually represent very small areas but have high temporal resolution, whereas the opposite is true for the latter. Therefore, the two types of measurements are complementary.
Frequency of measurement:
10 min. means at centres or experimental sites, hourly means at stations, monthly means from space.
Spatial resolution:
The measurements of albedo should be taken by hemispheric sensors mounted on towers as high as possible; the measurement locations should be chosen within relatively homogeneous regions. As a result, the spatial resolution of the surface observations can vary. The spatial resolution of satellite-based albedo depends on the field of view of the satellite sensors, ranging from 1 km for operational meteorological satellite to tens of kilometers for meteorological research satellites. These data can be degraded to the larger grids required by models (typically 250 km for GCMs at the present).
Accuracy/precision required:
Relative accuracy ± 10%;
Absolute accuracy better than ± 5%.
R and D needed:
Continue development of satellite-based inversion algorithms with the focus of retrieving spectral surface albedos; improve the algorithms for cloud detection, angular, spectral and diurnal corrections; synergistic use of ground and spaced-based data to provide a global albedo data set of high spatial and temporal resolution.
Present status:
Current global data sets on surface albedo have a relative discrepancy of 20-30% in low- and mid-latitudes and over 50% in polar regions (Li et al., 1997). There exists essentially two types of global surface albedo data sets. One is obtained by combining global cover type information with the typical albedo values obtained for each cover type (Wilson and Henderson-Sellers, 1985). Such data sets generally do not contain sufficient temporal resolution and lack information on the interannual variability. The other is derived from clear-sky satellite measurements (Staylor, et al., 1990; Li and Garand, 1994), which are hindered by the presence of clouds and insufficient knowledge of the atmospheric state. The major limitation for both types of data is the lack of spectral variation that is required by most models.
Measurement method:
Tiers 2-3: tower-mounted hemispheric sensors;
Tier 5: satellite measurements.
Associated measurements:
Upwelling and downwelling shortwave radiative fluxes at the surface; shortwave radiative flux at the top of the atmosphere, sky condition, and atmospheric parameters (water vapour, aerosol).
Action required:
Promote the installation of towers mounted with downward- and upward-looking, hemispheric shortwave radiometers;
Promote the acquisition of global aerosol data sets.
RADIATION - FRACTION OF PHOTOSYNTHETICALLY ACTIVE RADIATION ABSORBED BY THE PLANT CANOPY
Units of measure:
Dimensionless.
Users:
Climate, ecosystem, BGC modellers.
Rationale:
FPAR is an essential parameter relating the available visible solar radiation to its absorption by chlorophyll for plant photosynthesis. It is one of the surface parameters quantifying the CO2 uptake by plants and release of water through evapotranspiration. Since FPAR can be inferred from satellite measurements of reflected radiation (Myneni and Williams, 1994; Sellers, et al., 1994;), it is used as an intermediate variable in estimating the total light absorption by green plants and thereby plant productivity (Goward and Dye, 1987; Prince, 1991a; Goward and Dye, 1995). Because of the seasonal vegetation dynamics, FPAR can change rapidly, especially for herbaceous, deciduous, or short-season vegetation types. FPAR can be measured with ground optical instruments (Chen and Cihlar, 1995a; Pinter, 1993). Given its variation with the solar zenith angle, multi-temporal measurements are required. For the retrieval of FPAR from satellite data, a minimum of two spectral bands (in red and near infrared regions of the electromagnetic spectrum) are necessary (Huete and Liu, 1994). Satellite methods require cloud-free optical data to estimate FPAR, thus compositing of the data to minimize cloud interference is required. Corrections for diurnal variations are required when using satellite data to estimate FPAR (Chen and Cihlar, 1995b).
Frequency of measurement:
Globally, daily measurements are needed to produce products at 10-day intervals. For ground measurements, the frequency will be lower but more precise measurements will be acquired, e.g., the fraction of coverage by various layers in the canopy (trees, shrubs, understory) and their relative proportions in the canopy.
Spatial resolution:
Ground -10 m - 100 m;
Satellite - IOS: 1-8 km for global products;
Satellite - Post-IOS: 0.25-1 km.
Accuracy/precision required:
10-20% of true value;
5-10% of true value.
R and D needed:
Validation of present methods in various biomes. Improvement of methods to increase robustness and to reduce dependence of ground data. Improvements in the canopy radiation models used as the basis for inferring FPAR from satellite measurements and for relating satellite values to various levels within the canopy.
Present status:
Definitive algorithms are not yet available. Substantial R and D on FPAR estimation from satellite optical data has been carried out through both modelling and experimentation. One global data set has been produced through the ISLSCP project (Sellers, et al., 1994). The product is based on 8 km AVHRR data, and its validation and use in models has been initiated. ISLSCP continues efforts to improve the methodology as do other research programmes, e.g., BOREAS, HAPEX-Sahel, etc.
Measurement method:
Tiers 1-3: field measurements;
Tier 5: inferred from satellite optical measurements.
Associated measurements:
Satellite data must be corrected for atmospheric effects, thus requiring additional information on the state of the atmosphere (especially water vapour, aerosols, ozone).
One of the following two parameters may be used in conjunction with FPAR to estimate the photosynthetically active radiation absorbed by green plants. One is the IPAR at the top of the plant canopy (Prince, 1991b). The other is the total photosynthetically active radiation absorbed by the vegetation (leaves, stem; live, dead)/soil complex (total APAR) (Li and Moreau, 1996; Moreau and Li, 1996). The latter approach is expected to yield estimates with a lower error because fewer assumptions about the atmospheric state are required.
Action required:
Testing and evaluation of existing products and methods. Testing the robustness of the present methods by developing data sets for several years;
Development of a higher resolution (1 km) data set using global 1 km AVHRR product (Eidenshink and Faundeen, 1994);
Validation of the APAR total approach mentioned above in various biomes. Continuing research (see above);
Transfer of methodologies to improved satellite data sources. Enhancement of product to take advantage of better data characteristics (radiometric, spectral, etc.);
Use of products in vegetation productivity modelling.
RADIATION - INCOMING SHORT-WAVE
Units of measure:
W/m2.
Users:
Meteorologists, climatologists, agronomists, hydrologists, cryologists, ecologists, energy planners.
Rationale:
Incoming solar radiation is the fundamental driver of hydrological processes and leaf and soil temperature. It is highly correlated with incoming photosynthetically active radiation, the fundamental driver of primary production.
Most of the incoming radiation (on an energy basis) is in the short-wave. The total short-wave instruments are simple, stable and widely distributed, and should be the station level standard.
The long-wave component can usually be inferred or modelled; it should be measured at the centre level to establish and validate these models. Similarly, the direct and indirect components can be modelled from cloud cover (and type) and atmospheric aerosol profile; centre level sites should be equipped with sun photometers capable of measuring these components independently, and by inversion, calculating aerosol profiles. However, since it is much easier to accurately measure total long-wave flux as well as shortwave direct and diffuse components these should be measured wherever possible.
Frequency of measurement:
Hourly means at station level; 10-minute means at centres. Derived global surfaces of monthly long-term means.
Spatial resolution:
Point measurements. The spatial representativeness depends on the temporal averaging used. Monthly mean fluxes may represent 102s km, daily means 10s km, instantaneous fluxes kms; all are subject to the effects of the cloud regime.
Accuracy/precision required:
5% absolute for total short wave and long wave, 2-3% absolute accuracy (5W/m2) for centre radiometers.
R and D needed:
None.
Present status:
Most agricultural and ecological research stations are equipped with short-wave radiometers; there are some problems maintaining calibration standards. There are existing networks of high-precision radiometers and photometers, which need expansion to be spatially representative.
Measurement method:
Tiers 1-3: direct measurements by radiometers;
Tier 5: inferred from satellite radiance measurements.
Associated measurements:
Cloud cover, aerosols, PAR.
Action required:
Establish inter-calibration protocols for short-wave radiometers, a spatially representative precision radiometer and sun photometer network, and routine satellite-based cloud and aerosol products adequate to support radiation modelling at any point on the terrestrial surface.
RADIATION - OUTGOING LONG-WAVE
Units of measure:
W/m2.
Users:
Hydrologists, climatologists, remote-sensors.
Rationale:
Outgoing long-wave radiation is an important component of the site surface energy balance. It can also be used to provide an integrated vegetation and land-surface temperature, which relates to, for instance, leaf physiology and stress levels in vegetation. By closing the site energy balance the evapotranspiration rate can be calculated or validated.
Frequency of measurement:
Hourly means at stations, 10-minute means at centres.
Spatial resolution:
The measurements of upwelling radiation are taken by hemispherical sensors mounted on towers; the area sampled depends on the height of the tower, which should be adjusted to gain a representative sample (a few m2 in grasslands, to a few hundred m2 in forests).
Accuracy/precision required:
10% absolute.
R and D needed:
Development of robust all-weather infra-red sensors.
Present status:
Not routinely collected except by satellites such as AVHRR.
Measurement method:
Tiers 1-3: field measurements;
Tier 5: inferred using atmospheric models.
Associated measurements:
Incoming short-wave, outgoing short-wave.
Action required:
Promote the use of infra-red down-looking sensors at sites with towers and short-wave sensors.
RADIATION - REFLECTED SHORT-WAVE
Units of measure:
W/m2.
Users:
Climatologists, hydrologists.
Rationale:
The net radiation is fundamental to energy (and water) balance at the land surface. It consists of the sum of the downwelling and upwelling radiation. The downwelling radiation has been described as a separate variable. The upwelling radiation consists of two components: the reflected short-wave radiation and the re-radiated long-wave. Measurement of the reflected solar radiation allows to determine a key climate parameter, surface albedo that is defined as the ratio of the downwelling to upwelling solar irradiance. Land surface albedo is subject to considerable spatial and temporal variations. Knowledge of surface albedo is not only critical to climate and ecosystem modelling, but also a sensitive indicator of land cover change. Climate modelling has suffered, for a long time, from the lack of accurate global surface albedo data (Dickinson, 1995). A recent survey (Li, et al, 1997) indicated that the relative discrepancy in surface albedo among the existing global data sets reaches 20% over snow-free land and 50% in snow/ice covered areas. The major problem in measuring reflected radiation is the poor spatial representation, which may be alleviated somewhat by selecting uniform and representative observation sites and deploy observation towers. However, it is not feasible to deploy extensively such observation facilities. Therefore, measurement of the reflected radiation must be aimed primarily at validating satellite-derived products. Acquisition of global uniform data on surface albedo can only be achieved by means of space-borne observation (Sellers, et al., 1995). As most models require wavelength resolved surface albedo data, it is highly recommended to measure the reflected flux at several spectral intervals, at least visible and near infrared, in addition to broad band measurements.
Frequency of measurement:
Hourly means at stations, 10-minute means at centres, monthly means for global terrestrial fields.
Spatial resolution:
The measurements of upwelling radiation are taken by hemispherical sensors mounted on towers; the area sampled depends on the height of the tower, which should be adjusted to gain a representative sample (a few m2 in grasslands, to a few hundred m2 in forests). Ideally, the resolution of surface measurements should be selected in relation to the pixel size of satellite data and the heterogeneity of land cover.
Accuracy/precision required:
5% absolute for total short wave and long wave, 2-3% absolute accuracy (5W/m2).
R and D needed:
Continue development of satellite-based inversion algorithms with the aim to retrieve accurate spectral surface albedos; improvements of the algorithms for cloud detection, angular, spectral and diurnal corrections; synergistic use of ground and spaced-based data to provide a global albedo data set of high spatial and temporal resolution.
Present status:
Upwelling radiation is not routinely collected except in a few detailed micrometeorological experiments. The standard instrument used is usually the net radiometer, which is troublesome for continuous use and does not resolve the upward and downward components.
Measurement method:
Tiers 1-3: direct measurements with radiometers;
Tier 5: inferred from satellite radiance measurements.
Associated measurements:
Incoming short-wave radiation, outgoing long-wave radiation.
Action required:
Promote the installation of downward-looking, hemispherical short-wave radiometers at sites which already have towers and upward-looking radiometers.
RAINFALL CHEMISTRY
Units of measure:
mg NO3/litre, mg SO4/litre.
Users:
Biogeochemical modellers.
Rationale:
Sulphate and nitrate in rainfall are one of the principal biogeochemical inputs of these elements to ecosystems, and are generally increasing as a result of increased emissions of SO2 and Nox. These are the principal components of 'acid rain'. They have direct impacts on ecosystem function, including enhanced productivity and carbon sequestration at low levels, and acidification and ultimately ecosystem failure at cumulatively high levels. The most widely used analytical technique is ion chromatography, which provides, with little additional cost and effort, values for all the major cations and anions. This is advantageous as it increases the understanding of the origins of rainfall contaminants and also permits an internal quality check, for balance of cations and anions.
Frequency of measurement:
Per rainfall event, accumulated (by freezing and/or addition of stabilising chemicals) weekly or monthly.
Spatial resolution:
One collector per site.
Accuracy/precision required:
5%.
R and D needed:
None.
Present status:
Widely but not consistently or continuously collected. Research networks exist (e.g., DEBITS under IGBP-IGAC).
Measurement method:
Tiers 1 to 3: direct measurement.
Associated measurements:
Rainfall amount per event, passive collectors for fry deposition.
Action required:
Coordinate with IGBP-IGAC DEBITS.
RELATIVE HUMIDITY (ATMOSPHERIC WATER CONTENT NEAR THE SURFACE)
Units of measure:
g H2O/m3.
Users:
Climatologists and climate modellers; ecologists and ecological modellers.
Rationale:
Relative humidity, or atmospheric moisture, controls potential evapotranspiration rate (see Evapotranspiration) from soils and vegetation surfaces. An increase in the intensity of the hydrological cycle predicted from GCM simulations of future climate would both increase relative humidity, and be modified by concommitant changes in atmospheric moisture. Because of its role in the linkage between vegetation and atmosphere (the end result of soil moisture being transported within and evapotranspired from vegetation surfaces), relative humidity is both a climate change forcing function, and a response variable. This provides it with a central importance in climate change.
Frequency of measurement:
For most purposes, in combination with evapotranspiration, wind and temperature measurements, daily or diurnal measurement of relative humidity is adequate. (Initially this will only be possible at tier 1, tier 2 and possibly some tier 3 sites and at existing weather stations where relative humidity is collected.)
Spatial resolution:
Points spaced at 10 to 100 km, including vertical humidity profiles. (This is the ideal, but it is recognized that initially the spatial resolution will be limited to tier 1, 2, and 3 sites, and at existing weather stations where it is routinely collected.)
Accuracy/precision required:
± 1%.
R and D needed:
Cheaper instruments that combine unattended remote recording and data logging by satellite transmission with solar power. Satellite imagery that measures atmospheric moisture is needed in the long term.
Present status:
Relative humidity measurement sites are erratically and sparsely distributed, mostly at airports and research stations. This network is entirely inadequate for modelling effects on the atmosphere, vegetation and hydrology, especially in mountainous terrain.
Measurement method:
Direct measurement.
Associated measurements:
Soil moisture, wind, and air temperature combine with relative humidity to define moisture transfer from vegetation to the atmosphere. Solar radiation, air and soil temperature, evapotranspiration, precipitation, vertically structured wind speed and humidity.
Action required:
In the short term, automated, data logging and transferring systems need to be distributed in places where relative humidity is rarely measured: areas of low population density, of high elevation, and in less-developed countries;
In the long term, instrumentation must be developed and placed to provide an automated collection of relative humidity data via satellite imagery.
ROOTING DEPTH - 95 %
Units of measure:
cm, m.
Users:
Climatic, ecosystem, BGC modellers and managers.
Rationale:
Data of rooting depth are important to properly quantify interactions between the climate, soil and plants, because the root growth is very sensitive to climate change and soil properties, especially moisture.
Frequency of measurement:
Every 5 years.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
Within 10% of true value.
R and D needed:
The main challenges are to find the correlation between rooting depth, soil and climate features.
Present status:
Generally speaking, the underground part of ecosystems are poorly studied. Data on rooting depth of different trees, shrubs and crops in different climate and soil conditions are not yet available.
Measurement method:
Tiers 1-4: field sampling (infer from soil and vegetation characteristics at tier 5).
Associated measurements:
Meteorological data and soil data are necessary. The growth rate and annual rate of turnover of roots should also be measured.
Action required:
Mapping of soil, climate and rooting depth for important species;
Standardisation of methods for measurement of rooting depth.
ROUGHNESS - Surface
Units of measure:
Metres.
Users:
Climate modellers who simulate vegetation-atmosphere water exchange; hydrologists who model vegetation-atmosphere water exchange.
Rationale:
Roughness is a vegetational structural variable and is important in models because it is used to calculate wind effects on evapotranspiration, and controls the rate of transfer of moisture from vegetation to atmosphere. Hence, GCMs require accurate estimates of surface roughness. For the same reason, hydrologic cycle models require accurate estimation of the vegetation-atmosphere linkage.
Frequency of measurement:
Same as land-cover measurements.
Spatial resolution:
1-10 m for vegetation surface roughness, measured at 1 to 10 km intervals.
Accuracy/precision required:
± 10%.
R and D needed:
Roughness is generally derivable from vegetation structure and associated variables. If techniques can be developed to measure directly roughness length, these should be pursued.
Present status:
Estimated qualitatively from qualitative descriptions of vegetation structure applied as an average value to vegetation units as large as biomes.
Measurement method:
Tiers 1-3: inferred from tower or aircraft flux measurements;
Tier 5: estimated using SVAT models for various canopy configurations (one option). Also can be related to cover type and topography, given that the relationships are properly quantified.
Associated measurements:
Vegetation structure and topography determines surface roughness. Related vegetation measures include above-ground and below-ground biomass, NEP, necromass, and vegetation structure.
Action required:
Roughness needs to be estimated at the level of continuous data fields and point data sets.
SEA ICE CONCENTRATION/EXTENT
Units of measure:
Concentration - Fractional coverage of total ice, km2;
Extent - Area within a specified minimum sea ice concentration (e.g., 15%) (km2).
Users:
GCM modellers, climate change detection analysts, climate impact modelling/analysts.
Rationale:
Changes in sea ice concentration (extent) play a major role in ocean-atmosphere fluxes of heat, moisture, and momentum. Climate models predict major changes in ice cover for a CO2 doubling with impacts on global climate and ocean thermohaline circulation. Ice concentration and extent are required to validate sea ice models and for change detection.
Frequency of measurement:
Daily.
Spatial resolution:
Concentration - 25 km;
Extent - 25 km.
Accuracy/precision required:
Concentration - cold season ± 1%, melt season ± 5%;
(Currently available 4-7% in cold season and 10% in melt season.)
Required Ice Extent position accuracy ± 25 km (this is currently available).
R and D needed:
Concentration and extent - Passive microwave mapping needs refinements for atmospheric effects, and algorithms for thin ice types need validating. Improvements are needed for SAR products in coastal areas and in marginal sea ice zones. Combination of active and passive microwave data.
Present status:
Multi-frequency passive microwave data operational since 1978. 25 km gridded and ice products (FYI, MYI total ice/open water) distributed on CD-ROMs by NSIDC (for SMMR and SSM/I). Radarsat Scansar and high resolution passive microwave data will permit mapping of leads and polynas. SAR adds only marginally to ice type mapping.
Measurement method:
Tiers 1-3: estimated from high resolution remote sensing images;
Tier 5: passive microwave measurements; spectral radiance measurements; microwave backscatter measurements.
Associated measurements:
Ice surface temperature, radiation components, snow depth on ice, melt pond fraction, sediment in ice.
Action required:
Maintain and intercalibrate SSM/I data stream to insure consistent and continuous long-term time series of sea ice extent;
Improve current retrievals to match required accuracy of ice concentration and extent.
SEA ICE MOTION
Units of measure:
km/h; direction (degrees of azimuth).
Users:
Climate and ocean modellers; shipping and offshore rig operators; operational and strategic planners.
Rationale:
Sea ice is a tracer of surface winds and ocean currents. Observational data are required to validate coupled climate models. Coupled atmosphere-ocean models predict sea ice concentration and drift based on thermodynamic and dynamic forcings. The predicted extent, concentration and thickness of ice depends on ice growth by freezing, ice drift by wind and currents, and ridging intensity. Model outputs of ice motion need validation by observed ice drift vectors.
Frequency of measurement:
Twice per day.
Spatial resolution:
10 × 10 km.
Accuracy/precision required:
Direction 10 deg.;
Speed 1 km h-1.
R and D needed:
Optimal blending of buoy, AVHRR, SAR, and SSM/I data on ice motion.
Present status:
Buoys - There are approximately 40 international Arctic drifting buoy platforms; and about 5 in the Antarctic; collecting position and pressure data (via the Data Collection and Position Location System (ARGOS)). AVHRR/ERS-1 SAR and SSM/I mapping research ongoing and test validation of sea ice models.
Measurement method:
Tiers 1-3: drifting buoys;
Tier 5: microwave temperature and backscatter (change detection from sequential images).
Associated measurements:
Pressure and air temperature from buoys.
Action required:
Maintain Arctic buoy network;
Expand Antarctic buoy network;
Improve associated temperature measurements;
Operational products from RADARSAT. (A geophysical processing system is being developed by NASA/Jet Propulsion Laboratory (JPL) for sea ice products and AVHRR Pathfinder (1982-1996.))
Improve techniques for extracting sea ice motion from satellite data sets (effort currently underway at NASA/Goddard Space Flight Center (GSFC) and AVHRR Pathfinder (1982-1996).
SEA ICE THICKNESS
Units of measure:
Metres (m).
Users:
GCM coupled modellers, climate change detection analysts, climate impact modelling/analysts.
Rationale:
Changes in sea ice thickness, especially the occurrence of thin ice, play a major role in modulating ocean-atmosphere fluxes. Climate models predict major changes in ice thickness for a CO2 doubling with impacts on global climate and ocean thermohaline circulation. Ice thickness is needed to validate sea ice models and for change detection.
Frequency of measurement:
Weekly.
Spatial resolution:
200 km.
Accuracy/precision required:
Required ice thickness distribution accuracy: ± 10%;
Currently available ice thickness accuracy for mean thickness: ± 15%.
R and D needed:
Processing of ULS data from submarines and stationary moorings needs to become more operational. Techniques for inversion and assimilation of satellite sea ice concentration measurements need to be developed for the purpose of obtaining ice thickness distributions. Some work is in progress.
Present status:
Sea ice thickness inferred from active and passive microwave sea ice typing (e.g., new, young, first-year, and multiyear). Very few ULS data from submarines in the Arctic have been released. These sensors actually measure ice draft below the sea surface. Some ULS measurements from moorings are available. There are also direct measurements from drilling holes in the Arctic and Antarctic.
Measurement method:
Tiers 1-3: direct measurements;
Tier 4: underwater profiling (e.g., sonar);
Tier 5: estimated based on ice type and seasonal ice climatology.
Associated measurements:
Ice surface temperature, radiation components, snow depth on ice, melt pond fraction.
Action required:
Develop and implement a strategy for extracting accurate sea ice thickness from submarine ULS records;
Augment moored ULS station network;
Process, archive and distribute submarine sonar data collected by the US Navy and the Royal Navy.
SNOW COVER AREA
Units of measure:
km2.
Users:
Modellers, climate, hydrology, climate impact analysts, change detection analysts.
Rationale:
Snow cover has major effects on surface albedo and energy balance. Snow albedo is 3 to 4 times higher than over other land surfaces (except ice). Snow melt is important in the seasonal change of energy flux and in soil moisture and ground water recharge. Mountain snow packs are a primary water resource in semi-arid areas (such as Central Asia and western North America) and a major economic base for tourism.
Frequency of measurement:
Daily.
Spatial resolution:
25 × 25 km.
Accuracy/precision required:
Area - nearest 25 x 25 km.
R and D needed:
Combining surface (station) measurements and satellite passive microwave and visible band data in an operational product (SSM/I, Advanced Microwave Scanning Radiometer (AMSR), MODIS). Research on canopy and other effects on passive microwave signatures. Evaluate/improve automatic SWE measurements where manned stations are being replaced.
Present status:
Weekly hemispheric snow extent since 1966/67 (NOAA/NESDIS);
Research - Status algorithms for passive microwave used routinely by AES Canada;
Basin-scale mapping in North America from the Geostationary Operational Environmental Satellite (GOES) and AVHRR (data on CD-ROMs) since 1990 (for hydrologic application).
Measurement method:
Tiers 1-5: spectral radiance.
Associated measurements:
Need vegetation canopy and digital elevation model information (one-time) to improve algorithms.
Action required:
Implement products containing optimal blend of satellite and surface data;
Assemble archive of station data from Eurasia and North America from beginning of records with associated metadata.
SNOW DEPTH
Units of measure:
cm.
Users:
Modellers, climate, hydrology, climate impact analysts, change detection analysts agriculturists, transportation specialists and recreational planners.
Rationale:
Snow cover has major effects on surface albedo and energy balance. Snow depth observations are important in assessing winter kill of agricultural crops, influencing surface energy exchange and managing snow removal.
Frequency of measurement:
Daily.
Spatial resolution:
Point measurements at existing climate and synoptic stations (GCOS surface network).
Accuracy/precision required:
± 2 cm at depths up to 20 cm;
± 10% of absolute at depths greater than 20 cm.
R and D needed:
Evaluate/improve automatic depth measurements. Need to define standards of observation for manual and automatic stations.
Present status:
Manual daily snow depth observations have been a standard observation at synoptic and climatic stations. Automatic snow depth sensor observations are being made at a few fully automated synoptic stations. There are experimental satellite derived estimates being made from microwave sensors.
Measurement method:
Tiers 1-4: direct field measurements (points, snow courses);
Tier 5: inferred from satellite data. The estimation based on optical data varies with cover type and is not generally reliable. Snow depth could be inferred using satellite microwave measurements if snow density distribution is known (see Snow Water Equivalent variable sheet).
Associated measurements:
Temperature and precipitation including snowfall.
Action required:
Implement automatic sensor observations;
Standardise manual observation methodology.
SNOW WATER EQUIVALENT (SWE)
Units of measure:
Liquid water equivalent (mm).
Users:
Modellers, climate, hydrology, climate impact analysts, change detection analysts, agriculturists, energy planners (hydropower).
Rationale:
Snow melt is important in the seasonal change of energy flux and in soil moisture and ground water re-charge. Mountain snow packs are a primary water resource in semi-arid areas (such as Central Asia and western North America) and a major economic base for tourism.
Frequency of measurement:
Daily (satellite derived);
Bi-weekly (snow courses);
Daily.
Spatial resolution:
25 × 25 km;
Snow courses representative of terrain and land cover within the vicinity of the GCOS surface network.
Accuracy/precision required:
± 20 % of absolute (satellite);
± 5 mm up to 100 mm;
± 10 mm for greater than 100 mm.
R and D needed:
Combining surface (station) measurements and satellite passive microwave and visible band data in an operational product (SSM/I, AMSR, MODIS). Research on canopy and other effects on passive microwave signatures. Evaluate/improve automatic SWE measurements where manned stations are being replaced.
Present status:
Currently there is no global satellite product being produced. The Canadian AES produces a weekly SWE satellite map for selected Canadian regions. Snow courses and pillow data are routinely collected and used by operational hydrologists.
Measurement method:
Tiers 1-3: field sampling; recording devices;
Tier 5: passive microwave emission (experimental).
Associated measurements:
Need vegetation canopy and digital elevation model information (one-time) to improve algorithms.
Action required:
Assemble archive of station data from Eurasia and North America from beginning of records with associated metadata;
Implement SWE products containing optimal blend of satellite and surface data.
SOIL BULK DENSITY
Units of measure:
Mg/m3 in a given depth range.
Users:
Hydrologists, biogeochemists, climatic modellers, agronomists.
Rationale:
Bulk density is a fundamental soil property which is both itself subject to anthropogenic impact, but is also essential to the interpretation of any nutrient budgets (for instance, to perform carbon inventories). Furthermore, it is used to calculate soil porosity, a crucial variable for hydrologists, climate modellers and trace gas modellers.
Frequency of measurement:
Decadal.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
5% absolute.
R and D needed:
None.
Present status:
Bulk density is measured widely but not systematically. As a result, a vast amount of soil data are rendered useless for global analysis because the elemental analyses are not accompanied by bulk density measurements. Bulk density is only routinely monitored at a few hundred agricultural and ecological research stations.
Measurement method:
Tiers 1-4: measured using field samples;
Tier 5: inferred from soil maps and management practices.
Associated measurements:
Bulk density is an essential accompaniment to soil C, N and P measurements. With soil C, it can be used to derive porosity.
Action required:
Establish bulk density as a routine measurement at point, station and centre levels as well as in global soil surveys.
SOIL CATION EXCHANGE CAPACITY
Units of measure:
cmol(+)/kg at pH 8.2 in 1M solution.
Users:
BGC modellers, agronomists, terrestrial ecosystem modellers.
Rationale:
The cation exchange capacity (CEC) is a measure of the soil's ability to retain cationic nutrients. It is also an index of the clay activity and mineralogy, which is important for calculating mineralisation rates, leaching rates and interaction with pollutants. It changes relatively slowly, and largely due to changes in the soil organic carbon and pH. It can either be measured at 'soil pH and ionic strength' or at a standard pH, such as 7 or 8.2, in strong ionic solution. The standard pH is selected here for the sake of universal applicability and comparability, since the CEC is to be used as an index of clay activity. The 'Effective Cation Capacity' in acid soils can be estimated from the sum of extractable bases plus acidity; these measurements are widely made and recommended in acid soils, but not obligatory on a global basis.
Frequency of measurement:
Once a decade.
Spatial Resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
5%.
R and D needed:
Cross-calibration with other methods at sites with long records using other methods.
Present status:
Widely applied, but not globally observed or standardised.
Measurement method:
Tiers 1 to 3: direct measurement;
Tier 5: inferred using soil maps.
Associated measurements:
pH, clay content, soil organic C.
Action required:
TBD.
SOIL MOISTURE
Units of measure:
Atmospheres of pressure (pascals) or cm/100 cm; both as available soil moisture capacity between wilting point and field capacity.
Users:
Ecologists using ET to estimate or model NPP, vegetation composition, vegetation structure; water and agricultural resource managers, hydrological modellers, atmospheric modellers estimating land-atmosphere water fluxes.
Rationale:
Of all measured climate variables, soil moisture is the most important determinant of NPP and vegetation structure, composition and density. The variable rate at which soil moisture is released back into the atmosphere by the vegetation results in a shift in time scales from that of precipitation, which occurs over a scale of minutes, to weeks. The partitioning of precipitation into soil moisture and run-off is highly dependent upon the holding capacity of the soil, the soil water content, the rain intensity, and the distribution in space and time of the rainfall. This results in another variable time scale, ranging from minutes to hours for a given rainfall pattern, to months and seasons for soil moisture.
Changes in soil moisture and evapotranspiration are likely to have large impacts on terrestrial vegetation, since the distribution and abundance of plant communities are controlled to a large extent by the quantity and seasonality of moisture. If changes in the water balance are significant, major shifts in vegetation patterns and conditions are a likely result of climate change.
Soil moisture is the integrator of surface and sub-surface water fluxes and is thus the key diagnostic variable for surface water budget. Many BGC processes are fundamentally related to soil moisture conditions. Hydrological models use soil moisture to determine runoff and evaporation rates.
Soil moisture has been shown to be a critical component of temperature and precipitation forecasts. These forecasts require soil moisture measurements down to 1 to 2 metres.
Frequency of measurement:
For data fields and point sets, weekly to annual; for stations, centres and surface experiments, daily maximum and minimum or diurnal (4 x daily) measurements.
Spatial resolution:
Tiers 1, 2 and 3.
Accuracy/precision required:
Area nearest 25 × 25 km grid;
± 5 mm up to 50 mm;
± 10 mm up to 600 mm.
R and D needed:
Radar imagery appears capable of measuring soil moisture, after several years of airborne and ground-based experiments, but it needs to be developed and instrumentation provided to satellite vehicles.
Present status:
There is no method to measure soil moisture directly on the needed time and space scales. Therefore, soil moisture is inferred from 4-dimensional data assimilation using models which have sufficiently sophisticated land-surface parameterizations. Runoff, evapotranspiration and precipitation are the most important soil moisture related variables in these models. Runoff and evapotranspiration are usually derived from models, but measured precipitation must be obtained. Despite the success of research in rainfall estimation through satellite techniques, precipitation estimates still use a combination of in situ and satellite data. Therefore the networks of rain gauges managed by weather services around the world are critical. It is evident that the actual density of these networks, despite the efforts of international agency programmes, is insufficient to provide global fields of precipitation for the needed precision in soil moisture estimates.
Measurement method:
Tiers 1-3: in situ recording sensors; site-specific water budget models;
Tier 5: microwave emission, microwave back scattter (experimental).
Associated measurements:
Soil moisture is a function of precipitation, and of soil texture, porosity and the rate of moisture withdrawal from the soil, that is, water uptake by plants and evapotranspiration from them, in turn controlled by stomatal opening, relative humidity and wind.
Action required:
Airborne and ground-based experiments and instrument development should proceed for eventual radar or SAR measurement of soil moisture;
Some current 4-dimensional data assimilation models are predicting soil moisture fields. Mechanisms are needed to make these data routinely and economically available;
Considerable improvement in the modelled soil moisture should result from a one-time effort to improve the specificity of global soil maps to include detailed physical properties and greater spatial detail;
Efforts should be made to support existing rainfall and runoff measurement networks, and to enhance their efficiency by using modern data collection, transmission, and dissemination systems.
SOIL PARTICLE SIZE DISTRIBUTION
Units of measure:
% soil mass in predefined particle size ranges.
Users:
Hydrologists, pedologists, agronomists, erosion scientists/sedimentologists.
Rationale:
Particle size distribution (PSD) is a fundamental soil property which has a strong influence on many other properties, and can be used to predict them. PSD (often referred to as 'texture', especially when classified into broad classes such as 'sandy' or 'clayey') is usually inferred, via Stoke's Law, from the sedimentation time of dispersed primary soil particles. It controls (with bulk density and soil carbon) the water retention and water transmission properties of the soil and the thermal capacity and conductance. The smallest particles (clays) are responsible for soil carbon stabilization and anion and cation retention. The physical cohesion of the soil, and thus its erodibility is related to PSD, soil carbon and chemical state.
In stony or gravelly soil, the volume fraction of stones (> 2 mm) needs to be recorded as well as the conventional analysis of PSD to 2 mm.
Frequency of measurement:
Decadal. (PSD changes over time, principally as a result of erosion, which selectively removes certain fixed fractions and soil layers.)
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
5% absolute.
R and D needed:
Research is needed on robust algorithms for inferring soil hydrological and other properties from PSD, and for converting between different national PSD class systems.
Present status:
PSD is one of the most widely-measured soil variables, but some computability problems remain, especially with respect to the particle class size limits. There are the rudiments of several global soil pedon databases.
Measurement method:
Tiers 1 to 3: direct measurement.
Associated measurements:
Most useful when collected in conjunction with bulk density and soil carbon. Stations and centres should also collect related hydrological functions (water retention curves and saturated hydraulic conductivity) and cation exchange capacity at pH 7.
Action required:
Standardise international class limits and develop methods to convert between national systems.
SOIL pH
Units of measure:
pH units, measured in 1:5 suspension in deionised water.
Users:
Biogeochemical modellers.
Rationale:
The pH of the soil is a fundamental property controlling soil biological and chemical processes, such as biological nitrogen fixation, root growth and the mineralisation of organic matter. It is also an indicator of soil acidification due to acid deposition resulting from industrial processes, or from agricultural activities. There are several standard ways of measuring soil pH, each with advantages under particular circumstances (for example, in 1M KCl, 0.01 M CaCl2, saturated paste extract). The proposal is for a single, easily-applied index.
Frequency of measurement:
Annual to decadal.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
0.1 pH units.
R and D needed:
Cross-calibrations to other pH methods for sites with long pH records.
Present status:
Widely applied, but not systematically observed globally and not standardised.
Measurement method:
Tiers 1 to 3: direct measurement,
Associated measurements:
Fertilizer use, rainfall, atmospheric deposition of nitrate and sulphate.
Action required:
TBD.
SOIL PHOSPHORUS - AVAILABLE
Units of measure:
mg P/kg.
Users:
BGC modellers, agronomists, ecological modellers, aquatic ecologists.
Rationale:
Only a small fraction of the total phosphorus in soils is available for uptake by plants. P availability is a major determinant of ecosystem productivity, particularly in the tropics, either directly or indirectly via its controlling influence on biological nitrogen fixation. P availability can change rapidly in agricultural systems, mainly due to fertilization and changes in the soil acidity. There is a wide variety of techniques for available P in soil, each with advantages in particular soil types. The proposal is for resin-extractable P, since this method mimics the availability to plant roots and is applicable to all soil types.
Frequency of measurement:
Once a year in agricultural systems, once a decade in natural systems.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
5%.
R and D needed:
Cross-calibrations with other standard available P methods for sites with long records.
Present status:
Available P is routinely measured in many locations by a wide variety of methods. The proposed method is well tested, but not universally applied, nor globally observed.
Measurement method:
Tiers 1-4: laboratory analysis of soil samples;
Tier 5: inferred from soil maps.
Associated measurements:
Soil bulk density for the same horizon, soil total phosphorus, fertilizer in use.
Action required:
TBD.
SOIL PHOSPHORUS - TOTAL
Units of measure:
g P/m2 to given depth (30 cm or 1 m).
Users:
Terrestrial ecosystem modellers, agronomists, aquatic ecologists.
Rationale:
After N, P is the most generally limiting nutrient in terrestrial natural and agricultural ecosystems, especially in the tropics. In many cases, it may be the ultimately limiting nutrient, due to its constraining influence on the N cycle. In freshwater aquatic systems it is the most generally limiting nutrient and the one most associated with eutrotrophication. In agricultural systems, it is the macronutrient most widely traded at a global scale and exhaustion of readily available stocks could limit global food production.
Soil P occurs in many forms, which have complex interrelationships. Total soil P is one of the few robust measures. There is no single, standardisable method for measuring 'available' P in all soils; there is nevertheless virtue in making this measurement, using an appropriate technique, at stations and centres. Full fractionation of soil P, into organic P, secondary P, extractable P and occluded P is appropriate at centres and stations, as is assessment of the type and degree of mycorrhizal infection. Characterization of the P-sorbtion relationship is appropriate for soils with high iron.
Soil P should be routinely measured to 30 cm, and to 1 m at stations and centres. Plant tissue P levels should be monitored annually, at senescence, at stations and centres.
Frequency of measurement:
Decadal.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
5% absolute.
R and D needed:
Relevance of fractionation.
Present status:
Soil P is the most under-represented variable in global soil databases. Total P is not routinely measured, although the techniques are well known. Extractable P is widely measured, but the techniques used are incompatible and likely to remain that way.
Measurement method:
Tiers 1-4: laboratory analysis of soil samples;
Tier 5: inferred from soil maps.
Associated measurements:
Soil bulk density for the same horizon, soil available phosphorus, fertilizer use.
Action required:
Establish total soil P as routine measurement at stations and centres and in global soil survey.
SOIL SURFACE STATE
Units of measure:
mm/s.
Users:
Hydrologists, agronomists, sedimentologists, desertification monitors.
Rationale:
The partitioning of water between infiltration and runoff occurs at the soil surface. The state of the top few millimetres is strongly influenced by land management; one of the major mechanisms of desertification, for instance, is a sealing of the soil surface as a result of vegetation denudation.
The process of infiltration is dynamically variable and highly variable spatially within the landscape. Therefore, at centre level (and possibly station level) the time course of infiltration needs to be measured and resolved into its two main components (sorbtivity and steady-state infiltration). The key measurement at station level is the infiltration rate under constant tension (disk infiltrometer), repeated at enough positions to be representative of the landscape facet. Large area experiments need to resolve the landscape level redistribution and surface ponding capacity as well. At point samples, a visual estimate of the percentage of the soil surface in different states (capped, crusted, litter-covered, stone covered) is a surrogate.
Frequency of measurement:
Annually at centres and stations, decadally at points.
Spatial resolution: Tiers 1, 2, 3 and 4.
Accuracy/precision required:
10% absolute.
R and D needed:
A standardised soil surface state classification and measurement protocol for implementation at points.
Present status:
Infiltration rates are presently only monitored in a few experimental situations. Recent technological innovations (e.g., the disk infiltrometer) have resolved many of the previous methodological issues.
Measurement method:
Tiers 1 to 4: direct measurement.
Associated measurements:
Slope, PSD, soil carbon, vegetation cover.
Action required:
Establish infiltrometry as a standard monitoring technique at stations and centres;
Develop and test a point characterization protocol.
SOIL TOTAL CARBON
Units of measure:
g C/m2 to a given depth below the surface.
Users:
Biogeochemists, hydrologists, agronomists.
Rationale:
Soil carbon is the largest terrestrial carbon pool and the largest biospheric carbon pool subject to anthropogenic impact at the time scale of decades to centuries. Knowledge of the size of the soil carbon pool and its dynamics is therefore essential for conducting a national carbon dioxide flux inventory and for modelling the global carbon cycle. Soil carbon is also a controlling variable for many other soil variables; e.g., water holding capacity, cation exchange capacity and erodability. Soil carbon stocks are one index of sustainable agriculture.
Total soil carbon consists of a number of fractions with different turnover times. At centres and stations the carbon should be fractionated into at least microbial carbon and particulate organic matter. The microbial biomass carbon should be measured more frequently; perhaps annually.
Most of the changes in soil carbon take place in the top 30 cm, but significant amounts of carbon lie at greater depths. Measurements at centres and stations should be 0-30 cm and 30 cm-1m.
Frequency of measurement:
Decadal; microbial biomass carbon annually.
Spatial resolution:
Tiers. 1, 2, 3 and 4.
Accuracy/precision required:
5% absolute.
R and D needed:
Research is needed on robust soil carbon fractionation methods.
Present status:
Soil carbon is routinely measured at most national agricultural research stations and ecological field stations and is a basic variable in soil survey. The techniques for measuring it are simple and require little equipment. However, it is only monitored in a systematic way at a few locations, mostly long-term experiments. Standardisation of sampling depth and inter-calibration is required.
Measurement method:
Tiers 1-4: field sampling, laboratory analysis.
Associated measurements:
Must be accompanied by bulk density for the same horizon. Significant synergy can be obtained by measuring total nitrogen and organic phosphorus on the same samples; process insight obtained by fractionation of microbial and particulate carbon and by measurements of 13C and 14C. Particle size distribution is closely correlated.
Action required:
Formalise and standardise soil organic carbon monitoring at stations and centres;
Initiate point-level global soil carbon survey; create a global pedon database as an interim measure.
SOIL TOTAL NITROGEN
Units of measure:
g N/m2 to given depth for soil; g N/kg DM for plant material.
Users:
Terrestrial ecosystem modellers, agronomists, trace gas modellers.
Rationale:
Nitrogen is the element most generally limiting primary production in natural and agricultural ecosystems. The capacity of ecosystems to sequester carbon, for instance, is constrained by the maximum ratios of carbon-to-nitrogen (C/N) in living tissues. Several important greenhouse gases (e.g., N2O and NOX and CH4) are tied to the soil N cycle. As with soil C, most changes occur in the top 30 cm, which should be the measurement standard, but stations and centres should measure to 1 m.
Total N analysis mostly measures organic N, which is not the form taken up by plants. However, inorganic N (NO3 and NH4) are very variable at short time scales and therefore not suitable for monitoring except at centres or in ground water and aquatic systems. The process of conversion of organic N to inorganic N (mineralization and nitrification) are suitable for measuring only at the centre level.
Plant total N differs between tissues, and must therefore be measured separately for woody tissues, leaf tissue and fine roots. In the latter two, it varies characteristically at seasonal scale; the key measurement is at the moment of senescence; the next priority is the maximum concentration. Centres may wish to monitor tissue N monthly. 5N is a naturally-abundant isotope which can be used for tracing the fate of N and elucidating the cycle.
Frequency of measurement:
Decadal.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
5% absolute.
R and D needed:
Rates of gross and net mineralization.
Present status:
The techniques for total N are well tried and routinely applied in hundreds of laboratories worldwide. Soil and plant total N is monitored at a few hundred agricultural and ecological research stations.
Measurement method:
Tiers 1-4: laboratory analysis of soil samples;
Tier 5: inferred from soil maps.
Associated measurements:
Must be accompanied by bulk density for the same horizons. Total C on the same samples should be measured. Fractionation into microbial N, organic and inorganic N and measurement of potential mineralization rates lend mechanistic insights. Correlation with soil NOX and N2O fluxes is useful information. In agricultural settings, it must be accompanied by records of N applied as fertilizer; at stations N deposition should be measured, and at centres denitrification, N fixation and leachate fluxes should be monitored. At stations and points, the occurrence of N fixing organisms should be monitored.
Action required:
Increase the number of centres and stations at which soil N is monitored;
Initiate a global soil and plant N survey, in conjunction with the global soil C survey;
Support global pedon and plant databases as an interim measure.
SPECTRAL VEGETATION GREENNESS INDEX
Units of measure:
Dimensionless.
Users:
Climate, ecosystem, BGC modellers.
Rationale:
The ratio between the surface radiance or reflectance of visible and near infrared radiation has been found to be related to a large number of vegetation properties including leaf area index. Underlying these empirical relationships is the fact that ratios between these bands are affected by the fraction of photosynthetically active radiation absorbed by the vegetation. Because water vapour transpires through the same stomata that allow CO2 to enter the leaf in global climate modelling, the ratio has also been found valuable in estimating the energy used in evapotranspiration.
Commonly used spectral indices include the simple visible to near infrared ratio (Red/Near Infra-Red (NIR)) and the normalized difference vegetation index (NIR-Red)/(NIR+Red). More complex ratios involving other spectral bands to reduce the effects of factors such as the soil background are under development.
Composites for ten day to monthly time periods of satellite images of these ratios are routinely derived to improve observations of the ground surface. Their value arises from the fact that by choosing the pixel with the maximum spectral vegetation index for each location, many of the obscuring effects of clouds and other atmospheric effects are substantially reduced.
Frequency of measurement:
Images of vegetation indices may be derived locally on a daily basis, but for continental to global scales, temporal composites (see rationale) over periods of ten days to one month are required. However, to generate the latter, daily imaging is needed to build up satisfactory images.
Spatial resolution:
Data with a resolution of 8 km is of considerable value in global modelling activities, but for other applications finer resolutions of 1 km are required and for detection of interannual change, resolutions of 250 m are desirable.
Accuracy/precision required:
In terms of the normalized difference vegetation index, which varies in absolute terms from -1 to +1 and for actual land surfaces varies from approximately 0.15 to 0.8, an accuracy of 0.05 in terms of the actual at-surface value is required and a goal of 0.02 is highly desirable. Achieving the latter will require the availability of considerable precision in atmospheric correction.
R and D needed:
Continuing efforts are required in the development of more advanced indices in which the effects of soil and litter background are reduced and in the development of indices in which bi-directional effects are considered.
Present status:
The AVHRR of the NOAA satellite system has been widely used to generate global products. It has a number of deficiencies because of orbital drift of the satellite, poor calibration, and much less than optimal band width especially of the near infrared band. Future systems, notably MODIS, the Vegetation instrument of SPOT IV and ASTER-2 of ERS-2 are expected to give much better observations.
Measurement method:
Tiers 1-3: field radiance measurements;
Tiers 4-5: remote sensing optical measurements.
Associated measurements:
Incoming PAR received at the surface is required so that the amount of energy absorbed by vegetation can be estimated; observations allowing for atmospheric correction are needed.
Action required:
The value of the long-term record from the NOAA satellites has been much reduced by drift of satellite orbits; and efforts must in future be made to ensure stability of overpass time of all satellites making observations for this index;
Daily products must be better navigated to reduce misregistration effects.
STOMATAL CONDUCTANCE - MAXIMUM
Units of measure: m mole H2O/m2/sec.
Users:
Biogeochemical modellers, GCM and NWP modellers.
Rationale:
The maximum stomatal conductance is a key parameter of the rate of carbon assimilation and transpiration from vegetation. It is needed for biogeochemical, hydrological and plant production modelling. It is anticipated that it will change as a consequence of elevated carbon dioxide and possibly changes in the climate. It is measured at saturating levels of sunlight, in the absence of water stress.
Frequency of measurement:
Once every decade.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
± 10%.
R and D needed:
None.
Present status:
A widely measured parameter with mature instrumentation, but not globally and routinely observed.
Measurement method:
Tiers 1-3: field measurements with standard technique;
Tier 5: estimated using optical remote-sensing measurements.
Associated measurements:
Leaf area, Leaf N, temperature and PAR at the time of measurement, species composition/cover type.
Action required:
TBD.
SURFACE WATER FLOW- DISCHARGE
Units of measure:
m3/s.
Users:
Discharge is a critically important variable for GCM modellers, climate impact modellers, modelling of extreme events, hydrological modellers, national and regional analysts/planners.
Rationale:
Surface discharge is a measure of the volume of water flowing through a river channel cross-section per unit time. It can also be expressed as the equivalent water depth per unit area, given the knowledge of the surface area contributing to the discharge. (Note that this latter form is sometimes called 'runoff; the term runoff is not used here to avoid confusion with runoff in hill-slope hydrology.) As one term in water budget equation, discharge is a very important quantity for GCM validation. Independently measured discharge can thus provide check on other parameters computed by the numerical climate model, e.g., surface evapotranspiration rate on the seasonal time scale. Freshwater discharge from river mouths also strongly influences oceanic circulation at interannual to decadal time scales. Discharge is a high priority observation because it is essential for climate impact modellers, for example models of climate change impacts on water resource availability and extreme events. Discharge from pristine river basins may be a fundamental climate change detection element since it naturally integrates over an area and is very sensitive to climate change and measurement errors are likely to be random.
Frequency of measurement:
Depends on the specific question being asked (see Spatial resolution). With regard to data archiving, for tier 1 experiments it is generally expected that 15 minute to 1 hour data will need to be collected. For tier 2, hourly to mean daily and for tier 3, mean daily to monthly. Data should be collected as specified by WMO.
Spatial resolution:
The spatial resolution of the data very much depends on the specific question being answered and on sampling tier;
Tier 1 (detailed experimental catchments): process-based studies generally conducted in catchments of up to 100 km2;
Tier 2 (small pristine catchments up to 1 000 km2): climate change detection. These catchments should have minimum interference from anthropogenic influence;
Tier 3 (catchments greater than 100 km2): used for national and regional runoff studies along the lines managed by the GRDC. Such catchments provide an overlap with tiers 1 and 2.
Note that there should be flexibility in the area specified for each tier depending on the specific question that is being answered and the environmental conditions, e.g., mountains vis-a-vis plains.
Accuracy/precision required:
The goal for all discharge data within ± 5% of true value;
Because of interfering factors, a ± 15-20% accuracy could be acceptable.
R and D needed:
Spatial and temporal aggregation and disaggregation for application of measured data for modellers' needs. Improvements in measurement technology appropriate for large rivers and rivers subject to ice jamming.
Present status:
Discharge measurements are routinely made by hydrological services (or other agencies), and easy to obtain. For the climate change assessment, observations from catchments with minimal anthropogenic impacts (benchmark, pristine), or with accountable impacts, should be used. A study is needed to ascribe a number of stations in various parts of the globe to document climate change and variability: density will not be uniform and depend on several factors, such as topography.
The quality of river runoff data collected by national, state, local services is very variable. Measurement accuracy is poor due to lack of reliable stage-discharge relationships (as logistic problems and lack of funding prevents a regular updating, especially at high stages), unknown changes in channel cross-section of different flood stages, and lack of current metre rating. Geographical coverage is deficient in many areas, especially Africa and Latin America (coincidentally where the major rivers of the globe are located). Overall, the data in the developing world are insufficient to meet national needs and to allow the global hydrological cycle to be closed. Moreover, in recent years the network has declined significantly in quality and quantity. WHYCOS can be the solution, but needs to achieve global coverage and solve the logistical problems of national hydrometric agencies in developing countries in a sustainable way.
Data are not often archived nationally or internationally in forms facilitating their long-term use. An effort to be commended and encouraged is that of GRDC in Koblenz (Germany) which is concerned with collection, archival and distribution of global river discharges, but geographical coverage and duration of the data sets have serious deficiencies.
It is immediately possible to specify 150 to 200 stations at tier 3 that are required by the GRDC to obtain regional and continental discharge to the oceans. These stations can become operational in the next two years. However, to define a very specific IOS, particularly for tiers 1 and 2, the climatologists and GTOS community need to define clearly the specific questions to be answered.
The network design needs to be carried out in conjunction with national agency needs. It is unrealistic to expect that individual nations will establish and operate stations exclusively for GTOS and GCOS. Therefore it is critical that sites be selected that serve multiple purposes. The selection processes for stations is different between the developing countries and the developed countries. In developed countries the most appropriate catchments from the existing set should be selected. In contrast, in most developing countries, it will be necessary to provide for upgrading and/or rehabilitation of stations to obtain a sufficient density.
There are a number of networks already in existence that will meet many of the climate needs of GTOS and GCOS. Many of the existing national networks are sufficient for climate purposes. Governments need to be encouraged to make their data readily available. It is suggested that GCOS and GTOS work through the FRIEND programme to encourage governments to make at least a subset of their data available to the GCOS/GTOS community. A study is needed to ascribe a number of stations in various parts of the globe to document climate change and variability. Density will not be uniform and depend on several factors, such as topography.
Measurement method:
Tiers 1 to 3: direct measurement.
Associated measurements:
River level, stage-discharge relationships, global digital map of river basin boundaries, time-dependent adjustment factors to river discharge data, accounting for withdrawal of water for human use (irrigation, municipal water supplies, etc.).
Action required:
For climate change detection, in addition to the 150 to 200 stations mentioned above, there is a need for additional stations at tier 3, located in homogeneous climates and with a minimum of anthropogenic influences (such as hydrologic structures). In the longer term, it is desirable to have at least one station for every 2° by 2° grid. The network design needs to be carried out in conjunction with national agency needs, since it is unlikely that individual nations will establish and operate stations exclusively for GTOS/GCOS. Therefore it is critical that sites be selected that serve multiple purposes. The selection processes for stations is different between the developing countries and the developed countries (provisions for upgrading and/or rehabilitation of stations will be needed in the latter case to obtain sufficient density).
Generally, it has to be assumed that the data provided are of sufficient quality, as there is so much variability in hydrological data that checking data quality after the fact is almost an impossibility.
The following enhancements are urgently needed:
Filling gaps in spatial and temporal resolution;
Upgrading of existing stations to improve collection, data quality and transmission of data;
Rehabilitating stations where the infrastructure has to be replaced;
Establish the macro-scale network proposed by GRDC and enhance it to obtain a homogeneous time series. These stations should be asked to update data to fill time gaps;
Improve transmission of data utilizing WHYCOS where possible;
The priority (in decreasing order) should be given to: filling temporal gaps in existing records; upgrading of stations; rehabilitation of existing stations; and installation of new stations to fill spatial gaps.
SURFACE WATER STORAGE FLUXES
Units of measure:
Areal extent measured in square metres; depth changes measured in metres; estimated volume fluxes measured in m3/observation period.
Users:
Hydrologic modellers; water resource planners; ecological modellers; climate modellers; limnologists.
Rationale:
Surface water storage occurs in lakes, reservoirs and wetland areas for water in its liquid phase. The volume of water in a surface storage unit at any one time is an integrator variable, reflecting both atmospheric (precipitation, evaporation-energy) and hydrologic (surface water recharge, discharge and ground water tables) conditions. Depending on the storage capacity of a reservoir, it may primarily reflect human control. However, if lakes and wetland areas are not being affected by excessive withdrawal, they are strongly driven by extant climate conditions and are important for assessing net climate effects over time. If climate change is occurring to a hotter and drier mode, then lakes and wetlands should reflect this promptly. Internally draining lakes are especially important.
Frequency of measurement:
Monthly.
Spatial resolution:
500 largest lakes (volume) and selected others.
Accuracy/precision required:
Depth changes/stage along estimated stage should be within 5%; areal extent should be within 5%; volume flux should be within 5%.
R and D needed:
Depending on the available information, area-volume and stage-volume relationships need to be developed for any lake or wetland used in this analysis.
Present status:
The availability of time series of lake area and volume is limited but not negligible. The availability of wetland areal coverage information is minimal and seasonal variation almost non-existent.
Measurement method:
Tiers 1-5: spectral radiance (high to very high resolution).
Associated measurements:
Depending on the available information, area-volume and stage-volume relationships need to be developed for any lake or wetland used in this analysis.
Action required:
Areal extent of surface water bodies should be estimated from satellite data and developed in monthly time series;
If appropriate volume-area relationships can be developed (confirmed in situ to 5%) then volume fluxes could be estimated. If not, in situ depth variations should be measured and depth flux-volume change relationships developed.
TEMPERATURE - AIR
Units of measure:
Degrees Celsius.
Users:
Climate modellers; climatologists; vegetation ecologists and physiologists; ecosystem modellers; decision-makers.
Rationale:
Temperature is the most commonly-measured climate variable, the climate variable most directly sensitive to greenhouse gas increases, and the most easily-documented indicator of climate change. Along with soil moisture, leaf and soil temperatures (for which air temperature is a reasonable proxy) define rates of plant growth and development, reproduction and establishment.
Frequency of measurement:
For some physiological processes of growth and seedling establishment, hourly mean and daily maximum and minimum temperatures are needed; for ecosystem and landscape-level processes of NPP, daily to monthly values have been used.
Spatial resolution:
When combined with a precise digital elevation model which differentiates multiple elevation values from single measurements of temperature, resolution of one point measurement every 1 to 10 km is adequate for many purposes. Physiological model development and validation may call for highly-concentrated local measurements.
Accuracy/precision required:
± 1 degree C.
R and D needed:
Cheaper instruments that combine unattended remote recording and data logging by satellite transmission with solar power would be useful in the short term. Air temperature in the lowermost atmosphere should be obtainable by remote satellite observation and thus related instrument development and testing is needed in the long term.
Present status:
Air temperature networks from ground-based observation sites are the best-documented climate variables. However, they are inadequate for current purposes.
Measurement method:
Tiers 1-4: thermometers.
Associated measurements:
This is one of a suite of weather variables often measured in the same place: solar radiation, air and soil temperature, precipitation, evapotranspiration, vertically structured wind speed and relative humidity.
Action required:
In the short term (next 5 years), ground-based measurement sites require supplementing, particularly in areas of low population density, rough topography, high elevation, and in less-developed countries, where few measurements are available;
In the long term, satellite remote-sensing instrumentation needs to be developed, tested and placed in service.
TOPOGRAPHY
Units of measure:
Metres/kilometres.
Users:
Ecosystem modellers concerned with land cover change and deforestation as it affects carbon cycle. Hydrological cycle modellers, including GEWEX and BAHC projects. Remote-sensing specialists carrying out high-quality image rectification for change-detection purposes.
Rationale:
Topographic data are necessary to properly quantify the interactions between the solar radiation, water, and the heterogeneous land surface, including the measurements of these interactions from remote platforms.
The most complex interaction the biosphere has with the climate system is through the elevational and orthographic effects of topography. Though topographic structure strongly influences the regional climate, which controls soil development, the water balance, and the distribution of vegetation, actual data on these interactions are sparse, and are generally limited to a few small experimental field sites. Air temperature and precipitation databases must be corrected for topography if they are to be of use to climate modellers.
The fact that existing precipitation measurements significantly underestimate input volumes, severely limits our ability to model the water balance over a large region, unless we can account for the orthographic effects of topography and improve estimates of precipitation volume.
The problem is also evident in the gridded global humidity data recently released by R. Leemans to go with his 1990 global temperature and precipitation database (Leemans and Cramer, 1990). Though the temperatures in the 1990 database were corrected for elevation, the humidities were not. Consequently, 30-50% of the grid points for some months exceeded the maximum saturation value for the grid point temperature. Not only must a database be elevation corrected, but because many of the climate parameters are coupled, the correction of all parameters must be done in a coordinated approach.
Frequency of measurement:
Once for a given spatial resolution.
Spatial resolution:
1 km for use with AVHRR-type data (and for regional applications);
10 m-30 m for high resolution data and for hydrological modelling.
Accuracy/precision required:
Vertical: 100 m with AVHRR data, 5 m - 20 m for work with higher resolution data and for hydrological modelling.
R and D needed:
None.
Present status:
In the near term, only topographic data presently available can be realistically expected to form the basis for global and regional data sets. The best quality global topographic data sets are presently classified and although discussions have been held for their declassification, this has not yet happened. The presently available Digital Chart of the World contains elevation information digitized from 1:1 million maps, with coarse vertical resolution. Most of the world's land mass has been mapped at 1:250 000, thus providing a good basis for digital elevation data with a grid spacing of 100 m. It should be noted that the vertical accuracy of the above data sets is likely to be marginal at best. For many local or regional applications, higher resolution data will be required, such as are available from 1:50 000 or higher scale topographic maps.
Measurement method:
Tiers 1-3: field surveys; GPS; aircraft laser altimetry;
Tier 5: satellite altimetry; topographic maps.
Associated measurements:
N/A.
Action required:
Work towards the release of the highest resolution global data set presently available. In parallel, evaluate the utility of DCW topographic data;
Pursue the availability of higher resolution topographic data (1:20 000 to 1:50 000) for regional to continental applications.
VEGETATION STRUCTURE
Units of measure:
Qualitative classes of structure (non-woody annuals and perennials; shrub land, savanna, open forest, closed forest); also as a continuous variable measured in mT/ha (see biomass).
Users:
Climate modellers predicting surface roughness; ecologists predicting changing vegetation carbon contents and NPP from successional state and absorption of PAR.
Rationale:
The structure of vegetation defines potential LAI and in turn, is determined by moisture balance and PAR. Hence, vegetation model validation requires the capability to successfully simulate vegetation structure as an intermediate product in projecting carbon status from physical environmental variables. At the same time, vegetation structure defines surface roughness potential (see Surface Roughness) and as such is a required component for development of climate and hydrological cycle models. Finally, vegetation structure is closely correlated with potential and realized above-ground biomass and, to some degree, below-ground biomass. Therefore, it may be used to estimate biomass stocks in certain situations where biomass measurements are impossible.
Frequency of measurement:
Annually to decadally.
Spatial resolution:
Tiers 1, 2, 3 and 4.
Accuracy/precision required:
± 5%.
R and D needed:
Vegetation structure can be determined by stereo pairs of aerial photographs at the scale of 1-10 m2; development of automated remote-sensing and analysis could prove to save considerable time and effort.
Present status:
Fairly well known or estimated at scales of 10 000 km2 but poorly known, especially in tropics, at finer scales. Close to completion of capability for automated measurement; scattered and inadequate estimates by subjective (educated) guess.
Measurement method:
Tiers 1 to 3: direct measurement.
Associated measurements:
Above-ground and below-ground biomass, NEP, necromass, vegetation roughness.
Action required:
Complete IOS measurement at level of fields and points; re-measure at 5-10 year intervals.
VOLCANIC SULPHATE AEROSOLS
Units of measure:
Mg S/volcanic source/month.
Users:
GCM, NWP modellers, GHG modellers, remote-sensing scientists.
Rationale:
The emission of sulphur from volcanic eruptions is the major natural source of sulphate aerosols, which have a profound impact on the global radiation balance. This is also a significant part of the global sulphur budget. The bulk of the emissions can be traced to a small number of major eruptions per year. Aerosol data are essential for accurate evaluations of satellite optical remote-sensing measurements over land.
Frequency of measurement:
Monthly totals per significant event.
Spatial resolution:
Point source.
Accuracy/precision required:
20%.
R and D needed:
None.
Present status:
Currently, national programmes monitor all significant eruptions. Ongoing monitoring takes place at selected active volcanoes.
Measurement method:
Direct measurement during events;
Inferred from satellite optical measurements (over ocean, accuracy variable).
Associated measurements:
Atmospheric aerosol thickness.
Action required:
Make contact with existing volcanology data centres.
WIND VELOCITY
Units of measure:
km/hr, m/sec.
Users:
Climate modellers, hydrological cycle modellers, ecosystem modellers.
Rationale:
Wind speed (along with available soil moisture, stomatal conductance, and the relative humidity profile and the boundary layer) controls evapotranspiration rate, i.e., vegetation-atmosphere coupling. Hence, modelling of the hydrological cycle, and understanding of vegetation-atmosphere coupling, require wind speed information.
Frequency of measurement:
Hourly maximum, minimum and mean is adequate for many climatic, hydrologic and ecological modelling purposes. Monthly and annual measurements useful in certain ecosystem (NPP) models as well.
Spatial resolution:
20 to 100 km, including vertical wind profiles.
Accuracy/precision required:
± 10%.
R and D needed:
Cheaper instruments that combine unattended remote recording and data logging by satellite transmission with solar power would be useful in the short term.
Present status:
Wind measurement sites are erratically and sparsely distributed, mostly at larger airports. The measurement networks are entirely inadequate for modelling effects on vegetation and hydrology, especially in mountainous terrain.
Measurement method:
Tiers 1 to 3: recording anemometers. Wind velocity estimates are also produced through 4DDA.
Associated measurements:
This is one of a suite of weather variables often measured in the same place: solar radiation, air and soil temperature, evapotranspiration, precipitation, vertically structured wind speed and relative humidity.
Action required:
Implementation of solar-powered meteorological instrument installations that log data via satellite to remote computers and that include wind speed and vertical profiles at the level of points and stations.
