Annex 4: Reports of the Ad Hoc Working Groups^[3]

The Chairman proposed three ad hoc Working Groups be established to select a small number of high priority variables to be included in preliminary analyses using the objective (colour code) method. Where possible, they were also invited to select three of utmost importance. Working groups were established to discuss atmosphere, ocean, and land surface variables.

1. Atmosphere Variables

Members of the Group included Dr Croom, Ms Gitonga, Dr Hinsman, Dr Karpov, Mr Mignogno, Mr Morgan, Mr Ratier, Dr Richter, and Dr Ryder.

The ad hoc Working Group discussed three general atmospheric issues:

(1) atmospheric dynamics; (2) global radiative properties; and (3) atmospheric composition. The variables were divided into primary and subsidiary measurements, together forming the list of 'core' parameters. The three parameters chosen out of the 'core' parameter list as most important ones are represented in the primary measurements.

(1) Atmosphere Dynamics

Primary measurements include:

Wind profile (for 3 levels)

Temperature profile (for 3 levels)

Humidity profile (for 3 levels)

Subsidiary measurements are:

Total liquid water content

Precipitation rate

(2) Global Radiative Properties

Primary measurements are:

Top of the Atmosphere (TOA) outgoing short-wave radiation

TOA outgoing longwave radiation

Cloud cover

Subsidiary measurements are:

Solar constant

(3) Atmospheric Composition

There are qualitative requirements, based on straightforward physical principles, and the existence of active research programmes in the field, which point to desirability of measuring:

Cloud composition

Aerosol distribution

Trace gas composition

However, at present these have not converged to quantitative requirements for long-term monitoring specified by the Atmospheric Observation Panel.

Regarding the issue of fluxes, Dr Ryder prepared a statement:

It would be highly desirable for the purposes of the GOSSP to monitor the fluxes of sensible and latent heat, and where significant, momentum across the boundaries between land, atmosphere, cryosphere and oceans. In general, such fluxes are not amenable to direct measurements from space. However relevant secondary measurements can be made from which at least some of the properties of such fluxes can be inferred. At the land, ice, atmospheric boundaries, such measurements can be characterised as time series of multi-spectral (visible, infrared and microwave) imagery, from which momentum of the ice edge, changes in albedo, emissivity, temperature, etc., can be inferred. These measurements are generally required at high spectral and spatial resolution, but low temporal (daily) resolution, provided that diurnal tidal effects can be dealt with.

2. Ocean Variables

Members of the ad hoc Working Group included Dr Desa, Dr Halpern, Mr Ishida, Mr Johannessen, Mr Lefebvre, Dr Spence, and Mr Withrow

The Working Group identified the following core ocean parameters:

Geoid

Ice Thickness

Ocean Colour (Biomass)

Ocean Surface Topography

Ocean Surface Wind Vector

Ocean Wave Spectra

Sea Surface Temperature (SST)

Sea Ice Concentration

Sea Ice Cover

Sea Ice Edge

Sea Surface Salinity

Discussions covered SST, ocean colour, surface wind vector, and various sea ice parameters. The feasibility of sea surface salinity was discussed and it was pointed out that recent efforts indicated that progress had been made recently in the measurement of this parameter. Ocean colour received a lot of interest. It was considered to be important because we do not know how to simulate it, or derive it from other observations. It was pointed out that this parameter could be even more useful when used in conjunction with other parameters such as SST. A system approach would be valuable and the relation between observations and models was critical. There was discussion on the users of the measurements and how the observations could be assembled under a coherent plan. The global coupled ecosystem model was identified as an important goal and eventually the modelling of the global environmental system.

The meaning of 'operational' was discussed, and it was agreed that while 30 days may be considered 'operational' in the ocean sense, there were many things such as algae blooms that required almost synoptic monitoring. SST, sea surface topography, and surface wind vectors were identified as being of importance in relation to long-time series climate parameters. Data relay capability, not highlighted in the first version of the space plan, was identified as an important requirement for future ocean observations. Measurements should require at least 2 Dual Swath Scatterometers or 2 Altimeters (at least one Topography Experiment (TOPEX)/Poseidon class).

The Working Group decided that it was not possible to select a sublist containing the three most important parameters out of the 'core' parameter list for the ocean. The Group believed that the parameters selected will depend heavily on the problem being addressed. A climate problem would bring out a certain sublist while a coastal 'problem' would require another set. In selecting high priority parameters, the crossover between land and ocean may need to be considered. The Group decided that it would be important to add an annex that described the progress made to those improvements. The need to take into account the socio-economic effects of the product is necessary to gain support for future instrument/product development.

3. Land Variables

Members of the Working Group included Dr Aschbacher, Dr Cihlar, Dr Kibby, Dr Missotten, Dr Mitchell, and Prof. Wingham

The Working Group adopted a series of six steps to review the lists of parameters:

(1) Start with variables identified by TOPC (except for sea ice) based on the GCOS/GTOS Plan for Terrestrial Climate-related Observations (GCOS-21) (as revised, pp. 60-61; see draft submitted to JSTC-VI) cross-checked with the list of “In Situ Observations for the Global Observing System” developed by Unninayar and Schiffer (1996), (Draft Paper available from JPO). In addition, the Group considered only land variables (i.e., they excluded variables related to atmosphere and terrestrial - atmosphere interactions) and the requirements in terms of biophysical information requirements (not in terms of quantities measured by satellite sensors).

(2) From among the variables from step (1), select those for which satellite data could in principle provide useful information (primarily in view of the limitations of the physics of electromagnetic sensing); (see columns 'Variable' and ' Step 3' below.)

(3) Carry out a cursory review of additional key variables likely to be needed by non-climate component of GTOS (land degradation, biodiversity, chemical pollution, water resources) for which satellite data can in principle provide information (see Step 2).

(4) From among variables that met steps (2) and (3), identify those where a good chance exists now of obtaining a useful long-term global data set with the use of satellite measurements. Assign a rating of 'H' = high, 'M' = medium, 'L' = low (includes variables modelled or inferred using other, more directly observed variables), or 'N' = none (may provide related information but cannot be presently produced to yield the variable as defined by TOPC).

(5) From among the remaining variables in step (4) select 10 to 12 priority variables that should be included in a satellite land mission.

(6) From among the remaining variables from step (5), identify three parameters, i.e., those that are critically important to GCOS/GTOS terrestrial objectives, where satellite-derived products will make an essential contribution, and where a long-term data set can be envisioned at this time. In other words, includes a judgement on the importance of the variable to GCOS/GTOS objectives.

Finally, the Working Group reviewed the specifications of the variables as given in the report “In Situ Observations for the Global Observing System” by Unninayar and Schiffer (1996).

The following table shows the list of variables that remained after the above steps:

Comments:

• The final choice of variables depends on the choice of the questions. The choices made here reflect mainly some aspects of climate change impact and feedbacks from the land surface to climate;

• Leaf Area Index (LAI) is a critical input for Biological Global Climate (BGC) modelling and General Circulation Model (GCM)/Numerical Weather Prediction (NWP) modelling;

• Net primary productivity is estimated using models with inputs of LAI, or models using other satellite-derived variables such as incident solar radiation, Fraction of Photosynthetically Active Radiation (FPAR), etc.;

• Surface roughness (aerodynamic) is presently derived with Soil-Vegetation-Atmosphere Transfer (SVAT) models;

• Net Ecosystem Productivion (NEP) is derived from Net Primary Production (NPP) and other variables using models;

• Biomass is detectable using remote sensing only at low levels (herbaceous and thin woody stands);

• Spectral Vegetation Index (SVI) is an intermediate parameter between a raw satellite measurement and a biophysical variable such as LAI and FPAR. In less sophisticated models (e.g., for NPP estimation) SVI is used directly. Although SVI is an absolutely essential parameter, it is not included in the final list to maintain consistency in the table (which emphasises biophysical variables, not satellite measurements);

• Stomatal conductance is a function of spectral radiance and SVI;

• Vegetation structure (physiognomy) is presently derived from the knowledge of vegetation type. Potential for direct estimation exists in future data (e.g., laser);

• Fire area is required to reduce uncertainties in the global carbon budget calculations (biomass burning), for use in atmospheric modelling (aerosols), and for trace gas modelling

• Land cover (including land cover change) is essential for many GCOS and GTOS objectives; for GTOS, higher spatial resolution will generally be needed;

• Land use is inferred from the knowledge of land cover;

• Soil moisture data cannot be obtained from satellites except for the near-surface layer (refer to GCOS-21 for discussion);

• Surface water storage fluxes require measurements of changes in both area (image-type data) and depth (altimeter-type data). Since most surface water over land is in small water bodies, such measurements are not presently feasible on a global basis;

• Ice sheet mass balance data are required to predict the contribution of ice sheets to sea level rise; its magnitude is the largest uncertainty in determining the cause of the present sea level change. Meeting this information need, implies measurement of the area extent (image-type data) for small ice bodies and volume (altimeter-type data) for both small and large ice bodies. This variable is not in the final list in order to keep the final list short. Although in situ measurements could in principle be used, this is not a practical solution given the vast areas involved;

• Snow Water Equivalent (SWE) is the amount of water within the snowpack per unit area. In principle, SWE can be estimated from passive microwave radiometer measurements. Present understanding does not allow useful estimates in forested areas;

• Snow cover area is important in the determination of cold season albedo, as an indicator of interannual climate variability, for vegetation production in some regions, and for reservoir management in some regions. The SWE distribution is also important; however, it is not presently feasible to produce accurate global data sets of SWE distribution. Snow depth data (reported on Global Telecommunications System (GTS)) are presently used as a surrogate of SWE. Thus it is suggested that if snow cover is required to improve knowledge of albedo, snow area rates 'H'; without it, it would not be in the sublist of three out of the ten 'core' parameters. Note that albedo can be measured directly from satellites, although it can be difficult to distinguish between snow and clouds;

• FPAR is a measure of the proportion of solar radiation between 400 nanometers and 700 nanometers, which is utilised in the process of photosynthesis. FPAR is a direct input into some vegetation models, and it can also be used to derive LAI estimates. As an information requirement, FPAR is rated lower than LAI because LAI has broader uses;

• Topography is essential for many applications, including the corrections of satellite data before use. It is therefore rated as 'H' in importance but it is not included in the final list because topography does not present a monitoring requirement;

• It was noted that the variable specifications given by “In Situ Observations for the Global Observing System”, by Unninayar and Schiffer (1996), do not always correspond to the TOPC specifications; this should be corrected.

Caveats:

The above analysis has been carried out as a quick exercise and it is essential that it be critically reviewed by specialists.

Importantly, requirements for GTOS non-climate objectives should be based on the defined information needs and variables for those objectives; these information needs should be specified by GTOS as a matter of priority.