The views and recommendations of the meeting regarding the various hydrological variables considered are recorded below. They relate principally to Annex 1 of the GCOS/GTOS Plan for Terrestrial Climate-related Observations, version 1.0, November 1995, GCOS-21, WMO/TD-No. 721, UNEP/EAP.TR/95.07 and a draft list of variables endorsed by the TOPC III at its meeting held in Cape Town, South Africa from 19 to 22 March 1996.
The number following the decimal point refers to the number of the question in Annex 7 For example, section 1.2 discusses whether the spatial and temporal resolution of precipitation data are appropriate.
1. Point Precipitation
1.1 The variables are correct, but it should be made clear that we are measuring total and solid precipitation and that liquid precipitation is obtained as the difference between the two.
1.2 The spatial density should be “as measured” for point precipitation. The frequency of derived values of cumulative precipitation should vary as follows:
Macro-scale studies: daily - but there is a real difficulty here because of the different times at which daily readings are taken around the globe (local time).
Meso-scale studies: hourly - with considerable precision being requested for some work, but probably unobtainable on a wide scale. One complicating factor is the micro-scale dynamics of certain rainfall events.
1.3 The raw data should be available, but they should be corrected for systematic errors before they are used. For both macro and micro scale studies: 0.1 mm should be the definition of precipitation as opposed to no-precipitation, with a desired accuracy of:
± 0.1 mm for £ 5 mm
± 2% for > 5 mm
Achievable operational accuracy: ± 5%.
1.4 We should make a concerted effort to use the data from existing sites. There are a very large number of these 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).
On a more general basis:
Macro-scale studies: the first need is for 40 000 stations based on an objective analysis. We can use existing knowledge of actual versus desired stations based on information available to GPCC, that will be published soon, 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.
1.5 It is important to note that:
GPCC currently receives only monthly data and no daily data. They make available metadata about the point data that they hold but they are presently not authorized to release the data themselves.
An IOS could well be based on the GPCC holdings and its plans for a 40 000 station coverage if the above daily data access restrictions were relaxed. There is still a need for additional data from some regions to get a good global coverage, and GPCC can easily identify these, in particular by country.
1.6 Actions
Pursue technological advances in the direct measurement of precipitation over oceans. [Directed at: WMO/GOOS]
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. [Directed at: GPCC]
Send a letter to all countries requesting that update of the precipitation measurements forwarded to GPCC, including the most recent data and all synoptic station data. [Directed at: GPCC, WMO]
Design a future data flow structure for precipitation data and products (end to end). [Directed at: DIMP, GPCC]
2. Gridded Precipitation
(Variable also referred to as “areal precipitation”)
2.1 These data will be derived from point (rain gauge) measurements, satellite (thermal infrared, passive microwave) imaging data on 2.5o x 2.5o grids, and digital radar measurements. The proposed variables (TOP1) are adequate.
2.2 As regards spatial resolution and frequency, we should distinguish between:
Macro-scale studies: 1o x 1o grid, with daily and monthly data at continental-scale
Meso/micro-scale studies: 0.01o x 0.01o grid, with hourly data from digital radar and raingauges.
2.3 For all scales, the aim should be ±-5% accuracy.
2.4 Here the same comments apply as for “point precipitation”.
2.5 An IOS can be based on GPCC data with the aim of obtaining a monthly 1o x 1o data product with global coverage, based on an internal (GPCC) compilation at 0.5o x 0.5o. These data will be from point measurements merged with satellite estimates.
Data products for continental-scale purposes should be made available at 0.5o x 0.5o grid cell size.
Data on 0.01o grids are or will be available from regional projects.
The gaps identified under “point precipitation” are also applicable here, but the coverage is better, albeit with reduced accuracy, because of satellite coverage.
The primary enhancements should be a decrease of GPCC cell size to 0.5o x 0.5o for the global products (strongly dependent on improvements in the spatial resolution of satellite products). From the viewpoint of data processing at GPCC there are no technical problems in decreasing the cell size, but in practice the unavoidable need for manual control and checking would impose a high burden on current GPCC staff. A 0.5o product should include accuracy/error index for each grid cell.
2.6 Actions
Include snow (water equivalent) observations from surface (possibly satellite) in gridded precipitation products. [Directed at: WCRP/GPCC]
Prepare a description of a potential 0.5o x 0.5o global monthly gridded precipitation product (making optimum use of the present data and providing confidence information) and obtain user evaluation of the product’s usefulness. [Directed at: GPCC, TOPC]
Subject to a positive assessment, implement routine production of 0.5o x 0.5o monthly precipitation products. [Directed at: GPCC]
Support the establishment of digital radar networks and associated surface networks. Archive derived products at meso- and micro-scale. [Directed at: WMO, regional agencies]
Continue improvements in precipitation estimation using satellites, especially improvements in spatial resolution of he final products.
3. Variable: Evapotranspiration
3.1 It was clarified that what is needed is actual and not potential evapotranspiration. Otherwise, the variables as specified were adequate, although the units might read “kg/m2/day” to bring them in line with those for precipitation.
As with precipitation, a distinction was made under most headings between point and gridded values.
3.2
Point: 10 - 100 m2 on a daily basis.
Gridded: 100 - 1000 km2 on a daily basis.
3.3 There exist no practical means of measuring evapotranspiration extensively (eddy correlation offers a means of measurements at few locations); most evapotranspiration (ET) data are therefore derived estimates. Unfortunately, there is no consistent approach and there are no broad-based data sets available at the present.
If the desire is only to monitor change over time, then absolute accuracy may not be important and consistency is the main requirement. However, if the data are to be used with precipitation and runoff in flux studies, then accuracy becomes important.
± 20% or better should be the desired accuracy.
3.4
Point: the aim should be to locate and use the data collected and/or available from various activities (Tiers 1 to 3).
Gridded: the value of current point data for producing gridded sets is highly variable, depending on the location of the individual sites.
3.5
Point: An IOS may be developed, provided Tiers 1 to 3 sites archive relevant data.
Tier 3 studies, some long-term, could provide measurements/estimates of various types, and these should also be archived.
Examples of the above are data from GEWEX, ISLSCP and IGBP studies.
It will be important to ensure that the evapotranspiration estimates produced by these projects are seen as having a wider (GCOS/GTOS) use and are developed and archived with this in mind.
Gridded: There may be considerable potential value in the estimates of evapotranspiration derived by SVAT models, in particular those models contained within GCMs run to reanalyze 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.
3.6 Action
3.6.1
Point: Evapotranspiration
Ensure that Tier 2 and 3 sites have quality ET data (through site selection, upgrade of measurements,..). [Directed at: GTOS, GOOS]
Ensure that ET data are archived and readily accessible. [Directed at: DIMP]
3.6.2
Gridded: Evapotranspiration
Prepare the description of a potential gridded ET product(s) prepared through data assimilation/reanalysis under some present or near future programs. [Directed at: WCRP/GPCC]
Assess the usefulness of such a gridded global ET product(s). [Directed at: TOPC, GTOS]
4. Snow Cover Area (SCA) and Snow Water Equivalent (SWE)
4.1 The variables currently proposed are adequate. Snow cover area is itself of interest in radiation studies, but as a hydrological variable it can only be used in conjunction with estimates of SWE. Hence the two variables are taken together.
4.2 The focus for SCA and SWE should be on gridded products. A variety of point data will be used as a basis for deriving the gridded values.
Macro-scale studies: the spatial and temporal resolutions should be equal to those given for precipitation, or better if the accuracy is acceptable at 1o x 1o or 0.5o x 0.5o grid scale.
Meso/micro-scale studies: as for macro-scale, but better if the accuracy is acceptable.
4.3 SWE accuracy could be less stringent than for precipitation.
4.4 For SWE, emphasis should be put on using existing sites, even though they are not of uniform density.
4.5 A useful IOS can be developed for SCA based on global data sets using:
synoptic reports from surface weather stations
SCA (as vectors/polygons): NOAA-AVHRR
SWE (gridded): SMMR, SSM/I (validation by in-situ data required)
re-analysis (GCM) studies: here the quality and adequacy of the product has yet to be established.
The gridded values for meso/micro-scale should be archived with information on the estimated accuracy.
Various enhancements are possible, especially in combining point and spatial measurements. ACSYS studies could act as a test bed.
4.6 Actions
Request snow data from synoptic meteorological stations to be supplied by nations to GPCC (see also 1. Point Precipitation, directed at: GPCC, WMO)
Compare gridded snow products derived from satellite data with those from reanalyzed meteorological data. [Directed at: CRP/WG on Numerical Experiments, Committee on Atmospheric Sciences, GPCC]
5. Freshwater Flux to Oceans
5.1 Volume per time unit.
5.2 Spatial - based on statement of requirements of GCM modellers, LOICZ and ocean community (GOOS); sensitivity analysis may be helpful.
Temporal - mean daily/monthly.
5.3 5% desirable, 20% tolerated. Application of advanced existing technology (e. g. moving boat, Doppler current meter) needed for measurement of large rivers.
5.4 To cover as much as possible of the land masses to integrate better the flux into the ocean. This is dependent on the total number of stations. No - a uniform density is not needed.
5.5 Yes - based on the GRDC project for assessment of freshwater inflow to oceans. Initial number of stations may follow the GRDC study on representation of river runoff from continents into the oceans. There are gaps in the temporal coverage of the data and there is a need to cover some regions of the world more completely. There is a need to search for existing stations closer to river mouths than stations available in GRDC database at present, yet still free of pronounced tidal effects.
5.6 Actions
Encourage countries to transmit to WMO homogeneous time series in a continuous mode. [Directed at: GRDC, WMO]
Upgrade and rehabilitate existing stations. [Directed at: GRDC, WMO]
6. Bio-Geo-Chemical (BGC) Material Transport to Oceans
(Variable also referred to as “material transport from land to oceans via rivers”)
Units of measure: concentration in mg (or micrograms, nanograms) per litre; fluxes in metric tonnes per year.
A longer-term objective should also be - monitoring pesticides and micropollutants.
6.1 Need for monitoring of other parameters: pH, conductivity, DO, BOD, water temperature, concentrations of silica and micropollutants. Assessment of loads and concentrations necessary.
6.2 Spatial: using the same tier 3 system as in the case of flux into the ocean. Attempt to co-locate discharge stations and water quality stations close to each other. Temporal resolution should be more variable: basically twice a month. Denser temporal sampling downstream of urban and/or industrial complexes. Event-based measurements with fine-time resolution data also necessary (e.g. inflow of agricultural pollutants). Refer to LOICZ.
6.3 Uniform accuracy for all variables concerned is impractical. Liaise with LOICZ.
6.4 Refer to LOICZ for criteria for site selection.
6.5 Yes based on the GEMS-Water project for assessment of BGC influx to oceans. There are a number of existing networks, the most appropriate of which is that of GEMS-Water. GEMS-Water does not have the capacity to cover all the major rivers. GEMS-Water lacks comprehensive country participation. Identify geographic areas where water quality sampling is deficient. Identify a national sampling systems that are not parts of GEMS-Water.
6.6 Actions
Support GEMS-Water, strengthen co-operation with LOICZ. [Directed at: GCOS/GTOS, WMO]
Upgrade and/or rehabilitate existing stations. Note: Many different methods exist for combining flow of water with chemistry and there are different ways of determining load. [Directed at: WMO]
7. Discharge
7.1 It is suggested that only one variable discharge be used, and that the term “runoff” be dropped. Discharge means volume of water flowing through a river for channel cross-section in unit time (after International Glossary of Hydrology, UNESCO and WMO, 1992). It can also be expressed as the equivalent depth of water per unit area. Runoff may be confused with the terminology associated with hill slope hydrology.
7.2 The spatial resolution of the data very much depends on the specific question that is being answered. It is felt that discharge measurements can be collected at three tiers. Tier 1 are detailed experimental catchments. These are process-based studies and are generally conducted in catchments of up to 100 km2. For tier 2 small pristine catchments up to 1000 km2 are appropriate for climate change detection. These catchments should have minimum interference from anthropogenic influence. Tier 3 catchments should be greater than 100 km2 and are used for national and regional runoff studies along the lines managed by the GRDC. Such catchments provide an overlap with tiers 1 and 2. The recommendation is 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-à-vis plains.
For temporal resolution again it depends on the specific question being asked. 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.
7.3 A specification of within 5% of the true value is extremely optimistic. In reality this level of accuracy may be achieved in tier 1 studies, and should be the goal of all data collection. In reality, due to tidal influence, ice dams etc. an accuracy of 15 to 20% could be considered acceptable.
7.4 No - a uniform density is not appropriate, and the specific sites to be selected depend very much on the question to be answered. In mountainous areas the density would need to be much greater than on a plain. There are two levels of site criteria. Technical criteria as specified by WMO, and criteria for a network design. The criteria specified in the report of the TOPC meeting in Cape Town seem appropriate. Specifically:
Each participating country should have at least one site per biome type;
They should be capable of providing the appropriate data;
Sites in under-represented systems have priority over already-represented systems;
Reasonable security of tenure is required.
All else being equal:
Sites where research is also carried out are preferred;
Long-established sites are preferred;
Existing sites are preferred over sites which need to be established;
National support for the site is preferred to sites dependent on external funding;
Accessible, practical sites are preferred.
7.5 Yes for some purposes an IOS can be defined. For example we can immediately specify 150 to 200 stations at tier 3 that are required by GRDC to obtain regional and continental runoff 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 following discussion provides general principles that are to be followed in selecting the IOS.
For climate change detection, in addition to the 150 to 200 stations mentioned in the paragraph above, there is a need for additional stations at tier 3, being located in homogenous climates and with a minimum of anthropogenic influences, such as hydrologic structures. In the longer term, it is recommended to have at least one station for every 2o by 2o grid.
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.
First, 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. There are really three categories of enhancements that need to be considered:
1) Filling gaps in spatial and temporal resolution;
2) Upgrading of existing stations to improve collection, data quality and transmission of data; and
3) Rehabilitating stations where the infrastructure has to be replaced.
The order of priority should first be given to filling temporal gaps in existing records; then upgrading of stations; rehabilitation of existing stations; and lastly installation of new stations to fill spatial gaps.
7.6 Actions
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. [Directed at: GCOS/GTOS]
Improve transmission of data utilizing WHYCOS where possible. [Directed at: GCOS/GTOS]
GCOS/GTOS working through UNESCO should request FRIEND for support in at least three areas:
1. Request at least a subset of the existing data sets be made available to GCOS/GTOS;
2. Advice on establishing the IOS and the selection of sites for tier 1 and tier 2;
3. Request the meta data catalogue.
[Directed at: GCOS/GTOS]
GCOS and GTOS secretariats working with HWR, and FRIEND should identify existing data rich, experimental catchments at tier 1, and do all possible to assure that these data collection programmes continue. For tiers 1 through 3 there is an urgency that GTOS/GCOS secretariats and HWR give the highest priority to working with national governments to maintain a minimum number of stations and to work with developing countries to obtain donor support for upgrading and rehabilitation of stations. [Directed at: GCOS/GTOS, WMO/HWR]
8. Surface Water Flux
(Variable also referred to as surface water storage fluxes.)
8.1 The variables are correct and no additional ones are needed.
8.2 The spatial and temporal resolution are appropriate.
8.3 The accuracy for water level should be + 1 cm to + 50 cm depending on the scale.
The areal extent should be + 5%.
Volume should be + 1% to + 10% depending on depth of water.
8.4 Spatial resolution of 500 lakes and wetlands is appropriate.
8.5
The selection can be achieved, but the operational system needs to be assessed. The IUCN inventory needs consulting. The European Environment Agency data base also needs consultation for the European sector. There should be a cross reference with ILEC.
8.6 Actions
Inventory of lakes and hydrometric networks. [Directed at: WMO/HWR]
The following changes are suggested to page 87.
Units of Measure
add - open water evaporation
change - “flux measured in m3/observation period” to “net change in estimated volume fluxes measured in m3/observation period”
Rationale
Replace sentence: “If climate change is occurring to a hotter and drier mode, then...”
To “If climatic change is occurring, then lakes...”
Accuracy/Precision
See 8.3 above.
R&D Needed
Assess wetland volume fluxes. The use of remote sensing for areal extent and improve lake level measurements using satellite altimetry.
9. Groundwater
Rather than answering the questions for this variable the meeting expressed considerable discomfort with the contents and context of the groundwater description as written. For example, there was recognition that groundwater withdrawals have led to subsidence and relative sea level rise in some local instances. The meeting saw the write-up as conveying a more pervasive global sea level influence that it felt was unfounded.
On the other hand, the meeting had a clear sense that shallow groundwater systems are important and are poorly documented components of the transport of radiatively active trace gases through the terrestrial environment, as well as indicators of the linkages between the hydrologic system and such climate conditions and events as the El Niño - Southern Oscillation (ENSO), at least for selected tropical and temperate regions.
The meeting strongly recommended completely rewriting the groundwater text to reflect the more direct climate-related uses as a basis for continued inclusion of groundwater as a variable in Annex 1 of the GCOS/GTOS Plan for Terrestrial Climate-related Observations.
Groundwater (Shallow Aquifer) Levels and Chemistry
Units of measure:
Depth to water in metres.
Concentrations in meq l-1.
Users:
Hydrologic and ecosystem modellers and researchers, climate system researchers.
Rationale:
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:
Monthly, and in conjunction with periodic storm events.
Spatial resolution:
10-100 observation wells at each process research watershed (Tier 1 sites - discharge).
Accuracy/precision required:
<1% of true depth.
Within 20% of true concentration.
R and D needed:
Improved understanding and modelling of ground water-surface water interactions, including water quality genesis, fate and transport of chemical weathering products, and solute transport of radiatively active trace gases.
Present status:
A relatively small number (<100), long-term process research watersheds currently exist world-wide.
Associated measurements:
Precipitation, discharge.
Action required:
Ensure continued viability and integrity of long-term, process research watersheds by continually demonstrating the value and utility of data and information produced at such sites to both research and resource management issues. [Direct at: WMO/HWR]
10. Lake and River Freeze-up/Break-up
10.1 Yes the variables correct, and no additional variables are needed.
10.2 Above 45o north and above 2,000 metres in elevation one river and one shallow lake with negligible anthropogenic impact should be selected for each 2.5o x 2.5o grid. Daily observations are required to identify time series of dates of freeze-up and ice cover break-up.
10.3 The accuracy is correct.
10.4 See section 10.2.
10.5 Yes and initial operational system is possible. (See page 53 of GCOS-21)
10.6 Action
A central archive and/or several regional archives (e.g. North America, Russia, Himalayas) are needed. [Directed at: WMO/HWR, GCOS/GTOS]
