GERB SCIENCE APPLICATION PLAN

(MSG-RAL-GE-PL-0009, Issue 3)

CONTENTS

1. Scope

2. Applicable Documents

3. List of Acronyms

4. Introduction

5. Discussion of Applications

6. Attribution of Applications to Particular Institutions

7. Data Requirements to support these Applications

8. References

9. Appendices

1. SCOPE

This document has two main purposes:

1. To identify the planned and potential applications of data from GERB and the users who will implement these applications in their research or operations.

2. To develop a general list of the types of data products which are required from GERB in order to support these applications. This list forms the basis for the detailed specifications required to specify the ground segment.

This document should be read in conjunction with the Science Case (AD 1), which provides the top level scientific justification for GERB.

The document attempts to summarise the applications identified by members of the GERB International Science Team (GIST), and provides appendices which summarise the application plans for each Institution.

2. APPLICABLE DOCUMENTS

The following documents are applicable to this plan. The author should be notified of any inconsistencies between the applicable documents in this plan or between the applicable documents themselves.

Ref	Document Number	Issue	Title
AD 1	MSG-ICL-GE-MS-0010	1	The Science Justification for GERB

3. LIST OF ACRONYMS

ATSR Along Track Scanning Radiometer
AATSR Advanced Along Track Scanning Radiometer
BRDF Bi-directional Reflectance Distribution Function
BSRN Baseline Surface Radiation Network
CAL/VAL Calibration and Validation
CERES Clouds and the Earth’s Radiant Energy System
CLRC Central Laboratory of the Research Council
DFBO Dipartimento di Fisica dell Universita’ di Bologna
ECMWF European Centre for Medium range Weather Forecasts
EOS Earth Observing System
ERBE Earth Radiation Budget Experiment
ESA European Space Agency
ESSC Environmental Systems Science Centre
EUMETSAT European Organisation for the Exploitation of Meteorological Satellites
FOV Field of View
GCOS Global Climate Observing System
GERB Geostationary Earth Radiation Budget
HIRS High resolution Infrared Sounder
IASI Infrared Atmospheric Sounding Interferometer
IRMB Institut Royal Meteorologique de Belgique
ISCCP International Satellite Cloud Climatology Project
LMD Laboratoire de Meteorologie Dynamique
MSG METEOSAT Second Generation
NWP Numerical Weather Prediction
OLR Outgoing Longwave Radiation
RAL Rutherford Appleton Laboratory
REFIR Radiation Explorer in the Far Infrared
ScaRaB Scanner for Radiation Budget
SEVIRI Spinning Enhanced Visible and Infrared Imager
TAFTS Tropospheric Airborne Fourier Transform Spectrometer
TRMM Tropical Rainfall Measuring Mission
UGAMP Universities Global Atmospheric Modelling Programme

4. INTRODUCTION

As discussed in the Science Case, the unique feature of GERB in comparison with previous measurements of the Earth's radiation budget is the very high temporal sampling afforded by geostationary orbit, albeit for a limited region of the globe. Many of the applications discussed in the following section exploit this new capability. Several applications also take advantage of the synergy between the data from GERB and SEVIRI, the operational imager on MSG. A notable example is the proposal from the Institut Royal Meteorologique Belgique (IRMB) to merge the two data streams to produce near real-time estimates of the radiation budget at the high spatial resolution of SEVIRI. There are also proposals to merge or analyse the data in conjunction with measurements from instruments on other satellites, such as ScaRaB, CERES and AATSR.

5. DISCUSSION OF APPLICATIONS

It is convenient to discuss the potential and planned applications of GERB data under a number of broad headings, recognising that there is bound to be significant overlap between some of the areas. This section provides only an overview of these applications, since the detail is contained in the plans for each individual institution given in the appendices.

5.1 Radiation Budget Studies

Studies of the Earth's radiation budget have been carried out using satellite data for over three decades. A wealth of climatological information has been archived and a number of important results obtained, notably from the Earth Radiation Budget Experiment (ERBE). GERB data are expected to contribute to this field in several important areas:

a) Radiation budget climatology over the region covered by MSG, including climatological means and temporal and spatial variability.

b) Impact of clouds on the radiation budget, including cloud feedbacks. This will require the separate archiving of clear-sky fluxes to enable calculation of the cloud radiative forcing diagnostic.

c) Water vapour greenhouse effect and the water vapour feedback.

d) Diurnal cycle of convection and land surface temperatures.

e) Retrieval of the surface radiation budget and comparisons with other satellite retrievals and with surface observations (e.g. from BSRN).

f) Combining a and e, it will be possible to study the atmospheric flux divergences.

g) Impact of aerosols on the radiation budget. Of particular interest in this region are wind-blown dust from the Sahara and other arid regions, aerosols from biomass burning over tropical Africa and sulphate and other pollution aerosols from populated areas, notably Europe.

For some of these applications (e.g. d) it may not be necessary to perform a

radiance-to-flux conversion, as the information required is contained in the calibrated radiances. However, it is expected that most applications will require this conversion, so that the results can be expressed in terms of fluxes. In addition, most applications will require correlative data on related variables to be most productive. For example, b, c and d all require information on water vapour and clouds. Such data are available from many sources, but perhaps the most relevant will be the retrievals derived from the SEVIRI data, since these will be available with temporal and spatial resolutions which are compatible with the GERB data.

5.2 Evaluation of Numerical Models

An important application for GERB data will be in the evaluation of numerical models, in particular the three-dimensional models used in climate simulations and numerical weather prediction (NWP). Since climate models are themselves being used in many of the applications identified in 5.1, that list of applications is also appropriate here. In testing the model simulations, two broad categories are also worth highlighting:

a) determination of the overall climatology of the model over the area covered by GERB data.

b) using GERB data to test the performance in particular regions, giving insight into the physical realism of the parametrization schemes and the interactions between them. This is expected to be a particularly fruitful application, since the excellent temporal resolution of GERB (15 minutes) and the close match with the timestep of a typical global model means that processes such as surface heating and convection can be studied on a timestep to timestep basis.

A quite distinct application involves the testing of forward models which are used for simulating the radiation budget from operational or other products (e.g. ISCCP, re-analyses).

5.3 Meteorological and other exploitation

Plans to exploit GERB data by the national meteorological agencies are still at an early stage of development. One application which is at a mature stage of planning, however, is the proposal by IRMB to merge the GERB and SEVIRI data streams in near real-time and to produce radiation budget products for limited geographical regions at the high spatial resolution provided by SEVIRI. Details of this proposal are available elsewhere. Other broad categories of potential applications include:

a) operational use by national/international Met. services

b) commercial: agriculture, solar energy (with regard to surface fluxes)

c) use by EUMETSAT to add value to SEVIRI data and products.

Discussions are in progress to develop these applications.

5.4 Earth Observation Science

This application is listed separately because it is expected that the GERB data will be widely used for research into the science of satellite remote sensing, as opposed to the climate and meteorological studies listed in section 5.1. Strong interactions between the communities is expected, however. Applications which are planned include:

a) analysis of GERB radiance data on their own (including CAL/VAL activities)

b) analysis of GERB radiance data in conjunction with data from SEVIRI

c) analysis of GERB radiance data in conjunction with co-located data from other satellite instruments (e.g. AATSR)

d) studies of the angular reflectance characteristics of clouds and surfaces

e) observing system applications ("GCOS"), including merging the GERB data with data from polar orbiting satellites (e.g. ScaRaB and CERES) to create well-sampled high level products

5.5 Education and Public Understanding

Apart from the research applications identified above, there are also plans to use the data in public awareness projects, such as those associated with millennium activities.

6. ATTRIBUTION OF APPLICATIONS TO PARTICULAR INSTITUTIONS

Table 1 identifies the institutions which are currently collaborating in the GERB project and the applications from the above list which they plan to implement.

¹ Near Real Time

7. DATA REQUIREMENTS TO SUPPORT THESE APPLICATIONS

TBD

Identify in general terms the types of products required to support these applications (e.g. level 1 radiances, level 3 products etc.), their temporal and spatial resolutions, timeliness and other essential characteristics.

8. REFERENCES

 TBD

9. APPENDICES: PLANS ARRANGED BY INSTITUTION

APPENDIX A: Imperial College, UK

The principal objective of this study is to validate GERB products developed through this programme by the comparison of GERB data with co-temporal and co-spatial aircraft measurements. The opportunity exists not only to better constrain models through improved data input, but also to develop better model radiative transfer algorithms within a given atmospheric vertical region. It is intended to use high spectral resolution instruments to provide a test both of GERB and model accuracy. This validation will increase the value of GERB to the users. Secondly, to make use of GERB measurements to improve climate change models via the provision of an accurate representation of atmospheric radiative fluxes at high temporal resolution.

Use will be made both of existing instruments, and one currently under development, the Tropospheric Airborne Fourier Transform Spectrometer (TAFTS) designed to measure net, upwelling and downwelling radiances over the spectral region 12-120 microns. Recent work has illustrated the importance of the far infra-red component of the water vapour spectrum (wavelengths greater than about 20 microns) both in influencing the present day climate, and in influencing how it may change in the future. Radiometric coverage across the complete infra-red spectrum (using further instruments such as the well validated HIS instrument) will also permit spectrally integrated longwave radiances to be calculated at given vertical levels, and areas of spectral overlap between instruments will provide checks on consistency of measurements. Comparisons between the spectrally integrated aircraft measurements and GERB observations will allow the calculation of the amount of trapping occurring between the given aircraft level and top of the atmosphere, providing further insight into the atmospheric vertical profile. In particular, direct comparison between GERB observations, and spectrally integrated TAFTS data will allow the fraction of the OLR located in the farinfra-red to be calculated.

Given the high spectral resolution available, the retrieval of atmospheric constituent and temperature profiles is also feasible. The procurement of observed profiles via co-temporal and co-spatial radiosonde and ozone sonde soundings will allow for the calculation of merged retrieved/observed profiles. Using these profiles as input to the radiative transfer algorithms described in numerical climate models, in conjunction with the GERB TOA longwave radiance as an upper boundary, the representivity of the modelled physical processes can be investigated. Again, radiance measurements taken at the tropopause level both for direct and retrieval purposes will allow for detailed investigation into the impact of stratospheric absorption upon OLR, with particular emphasis on the role of stratospheric ozone and water vapour,both areas of considerable current research.

APPENDIX B: University of Leicester, UK

Angular dependence of the planetary radiation:

Data from the ATSR series of instruments will be used to provide additional angular information on top of atmosphere radiances which will be applied to the interpretation of GERB data, which are always acquired from the same viewing angles. This is important for developing more representative albedo models for GERB.

The ATSR data will also be used to investigate the long-term stability of GERB's onboard calibration system, in conjunction with a continuous assessment of the long-term performance of the GERB detector system.

Surface contributions to the planetary albedo:

Almost the entire Atlantic ocean and about one-third of the Indian ocean constitute a significant fraction of the GERB scene (cloud-free part) and the Sahara and Sahel regions are clearly visible. Surface fluxes in these regions are dominant factors in any attempt to understand and break down the global radiation budget. Estimates of these fluxes and of their effects on the planetary radiation budget will be made and refined using techniques being developed by other GERB partners, involving dynamic models of the atmosphere, land and ocean, satellite data (from ATSR), in situ measurements and meteorological analysis fields. Over land, the unique time-sampling abilities of GERB will provide insights into thermal inertia and hence soil moisture.

Shortwave radiative properties of stratiform cloud:

For the study of the shortwave radiative properties of stratiform cloud, GERB offers the ability to observe the 0.3 to 4 micron radiance of cloud as the solar zenith angle increments in steps of less than 4 degrees. It also averages over a region large compared with high spatial frequency cloud inhomogeneities, but small compared with the horizontal extent of stratiform cloud masses. It is therefore proposed to investigate the dependence of reflected radiance on the local solar zenith angle, providing insight into the role of clouds in the radiation budget.

Public understanding/education:

A consortium in Leicester has received funding from the Millenium Commission to set up a National Space Science Centre (NSSC).

APPENDIX C: Hadley Centre, Meteorological Office, UK

The work will be carried out in the Unified Model Parametrizations section of the Climate Research and Development division of the UK Meteorological Office. The section is located in the Hadley Centre for Climate Prediction and Research and is responsible for developing the parametrizations of physical processes in all versions of the Met. Office's Unified Model. This model is used for both climate research and weather forecasting, with a variety of horizontal resolutions ranging from those used for climate prediction (about 250km) and global forecasts (about 90km) to mesoscale forecasts over the UK (about 15km). Significant improvements in resolution will be implemented in the near future.

GERB data will be used to evaluate the performance of the model over the region covered by MSG. We plan to include code in the model to use the fields to simulate the radiances and other products derived from GERB. Apart from assessing the radiation budget climatology of the model for this region, the excellent temporal sampling of GERB will enable us to test the simulation of the Earth's radiation budget on timescales hitherto unavailable. The 15 minute repetition rate of the GERB data is similar to the frequency with which the physical parametrizations are called in the global models (20-30minutes), allowing detailed assessments of the simulated diurnal cycles ofconvection and land surface temperatures and of the life-cycle of convective and synoptic disturbances. Such comparisons will be crucial in evaluating the physical realism of the parametrizations. The synergy with the data from SEVIRI will also be exploited. SEVIRI products will include retrievals of upper tropospheric water vapour and of cloud amounts and heights. Combined with the independent measurements of the radiation budget provided by GERB, these products will facilitate detailed comparisons of both the clear-sky greenhouse effect and of cloud radiative forcing. These studies will providevaluable information on whether the simulated water vapour and cloudfeedbacks are realistic.

It is anticipated that dedicated integrations of the Unified Model will be required for at least some of these comparisons, because the diagnostics required are not normally stored at high temporal resolution. We estimate that we will need the equivalent of a 10-year integration at the current global forecast resolution (90km). As with the temporal resolution, the horizontal resolution of this version is well matched to that of the GERBdata (about 50km at the sub-satellite point). Integrations with the lower resolution currently employed in the climate model will also be performed.

Specific investigations are planned using the operational NWP model. Comparisons with GERB data will enable an independent assessment of the quality of the initial model analyses and of the evolution of the forecasts. These comparisons will employ a case study approach, using re-runs of theoperational forecasts. The possibility of carrying out near real-time assessments is under consideration. The excellent temporal sampling provided by GERB will enable such comparisons to be made timestep by timestep through the forecasts: only geostationary data can be used to do this.

It is also planned to simulate the radiances observed by SEVIRI, to provide correlative information on the variables contributing to the radiances, notably the temperature, water vapour and cloud distributions. It is precisely because the GERB data are not used directly in the forecast model that they provide such a important source of validation. The GERB broad-band radiances and fluxes are controlled by the atmospheric temperatures, humidities, cloud cover and the surface temperatures. By comparing the clear-sky and all-sky observed and modelled fluxes we will extract quantitative information on the quality of these fields in the analyses and forecasts which will be used to make recommendations on how to improve the formulation of the model.

APPENDIX D: University of Reading, UK

Keith P Shine

Department of Meteorology

University of Reading

Research in the Department of Meteorology includes a wide range of areas which would benefit from GERB data. These include radiation budget and radiation forcing studies, studies of tropical convection and tropical weather systems and the use of remotely sensed data in studies of African rainfall and hydrology. Associated with this work are various numerical models ranging from cloud resolving models to coupled ocean-atmosphere models for which GERB data could provide important information for model evaluation.

Concerning data requirements:

1. Data Type: It is anticipated that most needs will be met by Level 3 flux data, although some of these needs may be met by calibrated radiance data, should this be available significantly earlier. If "clear-sky" products were to be available, these would certainly be used.

2. Timeliness: There is no specific timeliness requirement

3. Data Volume: Some needs will be met by using monthly-mean data with a diurnal resolution of, say, 1 hour. However, for short periods (say, individual, occasional months) and over specific regions, data at high temporal and horizontal resolution would be required for process and averaging studies. In the longer term, specific events are likely to become of interest, so ready access to archived data will become a requirement.

Applications:

1. Radiation Budget Studies

Work at Reading includes considerable activity in calculations of radiative forcing due to changes in the concentrations of greenhouse gases and aerosols.

GERB is anticipated to contribute to this work by:

- providing a detailed radiation budget climatology of the GERB region for comparison with model calculations and to help relate the observed variations to temperature, water vapour and cloudiness variations;

- facilitating the study of the impact of diurnal and spatial variations in radiation budget parameters on radiative forcing simulations which currently use relatively coarse horizontal and temporal resolution.

- more speculatively, GERB will observe regions with large tropospheric aerosol loading due to both Saharan dust and, more seasonally, biomass burning. It is possible that the signatures of these aerosols may be detectable by examination of fluxes in cloud free regions.

2. Evaluation of numerical models

Both cloud resolving models and general circulation models require data to help evaluate their performance. Particular areas in which GERB data at good time resolution is expected to contribute are:

- evaluation of model convection and cloud prediction schemes

- evaluation of surface process schemes by monitoring clear sky emission, over both land and ocean surfaces.

APPENDIX E: Royal Meteorological Institute, Belgium

The science requirements for GERB require that radiances be measured to a certain accuracy, and the basic GERB data processing system will provide this. In addition, there is a user need to derive top of the atmosphere short wave, longwave and total fluxes from the data. To do this as accurately as possible requires the accurately geo-located GERB data plus data from the SEVIRI instrument on MSG in order to identify the type of scene being observed (ocean, land, cloud) and the fraction of each scene type in each GERB pixel. As part of this programme, software will be developed to produce these flux products.

The data processing system will be structured so that the flux products will be delivered to the partner running climate models within 4 hours of the data being taken. This will allow the information to be regularly used and the improvement that they provide in the model output to be fully assessed. While the intention is to provide a continuous stream of data to the user community, it is not intended that the system be considered an operational one. In particular, there will be no redundancy built into the processing system, so that hardware failures will cause a delay in the delivery of products to the users. In order to make maximum use of the flux products, some of them will also be made available for general use on the Internet.

We will study the radiation budget climatology for Belgium, inclusive of surface observations. In the context of aerology and more specifically atmospheric energetics, we want to study the temporal relation and interaction of the terms of the primitive energy budget equation integrated over different space/time scales. This includes the determination of space/time integrated radiation divergence as well the usage of the different atmospheric variables, temperature, pressure, concentrations of watervapour, liquid water, ice at relative high time 3D space resolution. These last observation variables will need to come from all available observations. The purpose is to correlate the findings tospecific weather patterns.

In the framework of the Aladin numerical model over Belgium, we will integrate these findings in order to improve the modelling and usage of the energy budget terms. The demonstration of the feasibility of this work should lead to the incorporation of improved methods for the operational use by the Meteorological services, including improvement of the forecasts of the available radiation at the ground, useful for the management of energyresources and for agriculture.

By using the high spatial resolution SEVIRI data together with the GERB information to generate improved spatial resolution ERB data we will be able to improve the shortwave and validate the longwave SEVIRI channels calibration.

The correlative use of SEVIRI and GERB data is our main activity to perform the different tasks described above. It is our intention to use the last proven angular reflectance characteristics of clouds and surface in our developments and to update these when better ones become available.

This activity is planned in view to homogenise the ERB observation data. It consists of merging in an objective and coherent fashion the observations from GERB/SEVIRI, CERES and ScaRaB.

Data requirements to support our application :

- Top of the atmosphere radiation budget terms at SEVIRI time/space resolution (derived by ourselves from GERB and SERIRI radiance's).

- SEVIRI radiance's at full resolution (to be collected by RMIB receiving station).

- GERB radiance's in near real time (by ISDN line RAL-RMIB ?).

- Aladin model variables over Belgium (calculated at RMIB).

- Surface measurements of radiation budget terms (performed operationally in Belgium by RMIB).

Specific notes with regard to operational applications:

GERB data will be used, in conjunction with ground based measurements of the radiative budget components at the surface in Belgium, to verify the values computed by the ALADIN - Belgium fine mesh limited area model. Measurements of the radiative fluxes will allow testing and validating the radiative transfer scheme. Clouds effects, water content, aerosol optical depth and amount of energy retained by the atmosphere will be studied using GERB data and some ground level measurements available for Belgium. One goal of the use of near real time GERB data for validation of NWP model calculation at RMIB is to demonstrate the operational usefulness of such data (not available from low orbiting satellite).

APPENDIX F: GKSS, Germany

It is intended to retrieve the surface and atmospheric radiation budget from combined GERB and SEVIRI data on regional scales within the field of view (FOV) of MSG.

This task includes the adaptation of a narrow to broad band conversion algorithm for the set of SEVIRI channels. Having such an algorithm developed on the basis of the high number of MSG channels allows its transfer to instruments onboard other operational satellites as for instance the polar orbiters. This link between GERB measurements and those taken by low polar orbiters will support studies on regional radiation budgets over the more polar regions outside the GERB FOV.

Broad band converted SEVIRI data will be used to derive the top of the atmosphere (TOA) Earth Radiation Budget (ERB) on regional scales comparable to those of mesoscale and also cloud resolving models. In parallel, the SEVIRI data will be used to produce estimates of basic cloud physical and optical properties, giving insight on cloud variability within collocated large GERB pixels (instantaneous field of views: IFOVs). The combined evaluation of low resolution GERB and high resolution SEVIRI ERB products together with information on the cloud properties permits to study uncertainties of ERB inversion algorithms introduced by scene identification and radiance to exitance conversion, if applied to inhomogeneous scenes as for instance broken cloud fields.

The cloud properties, as derived from SEVIRI data, together with information on the atmospheric state, as taken for instance from the ECMWF analysis, will be used to retrieve the vertical atmospheric profile of shortwave (SW, solar wavelength) and longwave (LW, thermal infrared wavelength) radiative fluxes. The TOA GERB radiation fluxes will be used as constraint on the radiation profiles, leading to an adjustment of atmospheric and cloud properties. For deriving energy flux profiles over the Arctic from satellite data, an adjoint model ADAM (Arctic satellite Data Assimilation Model) was developed at GKSS in cooperation with the University of Hamburg. In addition, it is planned to adapt this model for use within the MSG FOV and to test its usefulness over this region.

In parallel, empirical estimates of the shortwave and longwave Surface Radiation Budget (SRB), based on the GERB data by applying for instance the Li-Leighton algorithm for the shortwave and the Gupta, et al. approach for the longwave, together with all available surface observations will be used for testing the applicability of the methods to derive vertical radiation flux profiles.

It is expected that the combination of the temporal high resolution GERB products with auxiliary SEVIRI data of high spatial resolution will lead to ERB products, which permit to study cloud radiation processes on regional scales. Such type of studies are of particular interest for modelers involved in continental scale regional projects in Europe or Africa.

It is further expected to develop and validate a method to determine the three-dimensional radiative flux divergence fields in the atmosphere. Such a new procedure of GERB data analysis permits to quantify the role of different cloud fields in relation to the three-dimensional distribution of the radiation fields. In addition, the proposed study will contribute to validate the GERB inversion algorithm with respect to scene identification and radiance to flux conversion. Further, the derived method for narrow to broad band conversion will link data of auxiliary low polar orbiters to the GERB measurements. Such a transfer to other satellite instruments allows one to study the radiation budget within regions outside the MSG-FOV or also for past time periods, when no ERB instrument was in orbit, but which are of particular interest for different modelling groups.

APPENDIX G: University of Bologna, Italy

Objective A: Improve GERB estimate of the far-infrared portion of the OLR. This objective makes use of complementarity with the REFIR project.

Method:

1. Simulate REFIR data in clear-sky conditions for a set of standard atmospheric conditions.

2. Simulate REFIR data for cloudy sky situations: development of appropriate cloud models and their radiative properties in order to simulate the effects of cloudiness in the REFIR spectral range.

3. Compare GERB simulations with REFIR simulations for a given set of atmospheric conditions to define algorithms to improve the far-infrared estimate obtained from GERB using REFIR data.

Objective B: Prepare for observational studies of the role of water vapour radiative feedback.

Method:

1. Perform detailed estimates of retrieval errors for middle and upper tropospheric humidity from simulated clear-sky REFIR data.

2. Perform detailed estimates of retrieval errors for middle and upper tropospheric humidity from simulated clear-sky IASI data.

Objective C: Compare simulated GERB OLR measurements with OLR estimates from polar orbiters.

Method:

1. Compute TOVS radiances, for a set of standard atmospheric conditions, including cloud effects.

2. Derive OLR estimates from TOVS, from radiances computed in C1, and compare them with GERB simulated values.

3. Derive OLR estimates from a subset of IASI channels.

APPENDIX H: University of Valencia, Spain (Dr Lopez-Baeza)

(a) The Remote Sensing Unit of the University of Valencia are at present trying to improve their methodology for the estimation of shortwave radiative fluxes by incorporating some of the operational meteorological products that are supplied by the National Meteorological Services (Melia, et al., 1997). They are currently studying some intermediate products obtained from a HIRLAM retrospective re-analysis for 23 June 1991. The assimilation of updated NWP model intermediate products into radiation flux estimation algorithms gives the Spanish group an opportunity to use more realistic information with a convenient frequency, so that we can have a physical basis to obtain cumulative values (daily, weekly, monthly, etc) out of a sparse number of satellite observations, and this can be used for monitoring purposes in a quasi-operational way. In this way, satellite estimations, which usually require long computing times for derived physical quantities, especially radiation fluxes, may be carried out quasi-operationally and use them for monitoring studies. In particular, we think that the application of these ideas to GERB data may be of interest for the GIST.

(b) The spectral narrow-band to broad-band transformation for albedo has been studied with particular interest and in detail by the Remote Sensing Unit of the University of Valencia (Valiente et al., 1993, 1995, Russell et al., 1997, Tristan, 1997 y Tristan et al., 1997) and that has permitted an estimation of surface albedo with a high degree of accuracy using METEOSAT data (Fortea et al., 1997). The METEOSAT Second Generation (MSG) satellite will also carry on board a multi-channel discrete band radiometer called SEVIRI which, with the same sampling time as GERB, will permit calculation of radiances and, therefore, radiative fluxes, with better spatial resolution at specific regions during specific times. The opportunity of having available simultaneous narrow-band (SEVIRI) and broad-band (GERB) measurements, from the same space platform, will allow us to check and optimize the narrow-band to broad-band spectral transformation algorithms.

(c) Parameterization of cloud cover and cloud type effects on net solar radiation. The Remote Sensing Unit of the University of Valencia are progressively gaining experience in this area and they can estimate cloud cover and classify different cloud types from satellite images, by using different procedures (Femenia, 1995; Lopez-Baeza et al., 1996; Fortea , Lopez-Baeza, 1997). Presently, they are analysing data from the SRB, ISCCP and ERBE Projects over the Mediterranean basin. The opportunity of having simultaneous observation of SEVIRI and GERB will be used to carry out some estimations with SEVIRI of the basic physical and optical properties of the clouds contained within GERB field of view. In addition, the combination of simultaneous products of clouds and radiative fluxes allows the development of conceptual models of cloud processes.

APPENDIX I: European Centre for Medium Range Weather Forecasts, UK

At ECMWF, comparison with satellite measurements of the radiation fields at the top of the atmosphere is an important part of the model validation at the time of development of a new or modified parametrization. Outgoing longwave radiation and absorbed shortwave radiation at the top of the atmosphere produced by 4-month long seasonal simulations with the T63 model are usually compared to observations from Nimbus 7 or ERBE on a monthly or seasonal basis. For shorter timescales, comparisons are made with ISCCP or HIRS radiances. Within the framework of the verification of a weather forecast model, ECMWF expect GERB data to provide the best of both worlds: true fluxes (at least spectrally, if not directionally speaking) providing a good sampling of the diurnal cycle in near-real time.

APPENDIX J: NASA Langley Research Center, USA (Dr G. Louis Smith)

Areas of mutual interest may be grouped under three headings:

1. Validation.

Unfiltered Radiances:

Once during each orbit, the radiances observed by the CERES are the radiances observed by the GERB. This gives several times each day for which the unfiltered radiances can be compared as a check on calibration and unfiltering for the 2 instruments when the CERES is in crosstrack scan mode. There are many more opportunities when the CERES is in alongtrack scan mode.

The geometry for this coincidence of radiances is described thus. For crosstrack scan mode, the CERES is scanning in the plane normal to its velocity vector. This crosstrack scan plane contains the line which is normal to the orbit plane as the spacecraft goes around the Earth. The crosstack scan plane may be thought of as rotating about the orbit normal. When the crosstrack scan plane intersects the Meteosat then there is a line within the plane from the GERB through the CERES to the Earth. When the CERES and GERB are aligned with this direction, they are both measuring the same unfiltered radiance. This geometry will apply for both the CERES aboard the TRMM spacecraft, which will be in a precessing orbit of 37 degree inclination, and the CERES aboard the EOS AM and PM platforms.

When the CERES operates in biaxial scan mode, there will even more opportunities for the line from the scene to CERES to intersect GERB. The CERES/TRMM will operate in biaxial scan mode every third day. There will be 2 CERES instruments on the EOS AM platform, one of which will operate in crosstrack scan mode and the other in biaxial scan mode. There will also be 2 CERES on the EOS PM platform operating similarly.

The CERES/TRMM will also operate in alongtrack scan mode every 15 days. In this case the scan is in the orbit plane. When the GERB crosses the orbit plane, one constructs a line from GERB through CERES to the Earth. As the CERES moves along its orbit, this line follows it. As it scans alongtrack, the CERES will measure the radiance along this line, as does GERB. There will be a large number of lines from the Meteosat through CERES to the Earth for which the same radiance is measured by each instrument.

BRDFs:

In order to compute flux from a single radiance, it is necessary to use a Bidirectional Reflectance Distribution Model. The error in BRDFs is the major error in the retrieval of instantaneous fluxes at the "Top of the Atmosphere." This problem is even more serious for a radiation instrument in geosynchronous orbit than for polar orbiting spacecraft, because the GERB always views a given site at a constant view zenith angle and for this site the azimuth angle between the GERB and the sun is uniquely related to the solar zenith angle. (To be correct we must separate morning and afternoon, else the relation is 2-valued.) For sun-synchronous spacecraft a given site will be viewed from severval directions, so that some cancellation of errors will occur in computing a monthly=mean albedo.

Once the unfiltered radiances have been validated, one can select locations which GERB and CERES are both viewing at nearly the same time. The level to which the fluxes agree for the 2 instruments measures the accuracy of the BRDFs. Given enough pairs of such measurements, if the accuracy is not adequate one can upgrade the BRDFs by use of the Radiance Pairs Method (Green, 1996?).

2. CERES provides upgraded BRDFs.

Every third day the CERES/TRMM instrument will operate in biaxial scan mode so as to gather data for development of improved BRDFs. The EOS AM and PM spacecraft will each carry 2 CERES instruments, one of which will scan in crosstrack mode and the other of which will scan in azimuth as well as nadir angle. These biaxially rotating measurements will be used to develop improved BDRFs, which will be used to improve CERES data products. These BRDFs can also be used for GERB.

There will be a set of ERBE-like BRDFs, which can be selected by an ERBE-like scene-identification algorithm, and a set of CERES BRDFs, which will require a high resolution scanner for scene identification. In order to use the latter set, GERB would use the SEVIRI data, which is included in the data processing stream at IRMB.

3. GERB provides time sampling for Meteosat region.

One of the major sources of error for the ERBE daily mean and monthly mean data products is time-sampling error. In oder to form these means, it is necessary to interpolate between the measurements in order to form an average. The same will be true of CERES. Current plans call for use of ISCCP results to provide ancillary data for use in time-interpolation.

GERB data will provide enough time samples of radiation measurements that the time-sampling error will be negligible over the portion of Earth which is viewed by GERB. A combined CERES/GERB data product could reduce BRDF errors and have negligible time-sampling error.

APPENDIX K: NASA Langley Research Center, USA

Studies of Diurnal Processes

M. G. Mlynczak, G. Louis Smith and Bruce A. Wielicki

GERB measurements will be useful for a broad range of radiation budget studies. The present discussion is limited to diurnal processes, for which GERB will provide accurate radiation measurements for the first time. Diurnal processes have been studied using Meteosat and GOES data from narrow band channels (Minnis and Harrison, 1984; Duvel and Kandel, 1985; Cheruy et al., 1991), but deficiences of extracting broadband radiation components from narrowband channels limit the accuracy with which one can retrieve outgoing longwave radiances and reflected solar radiation. Diurnal variations are strongest in low latitudes, where radiation plays a major role in deserts and in deep convective processes. A better understanding of weather and climate in these regions will be provided by GERB data. Because weather systems are quite mobile, weather in these regions can ultimately affect other regions far away. We discuss some phenomena for which GERB will provide greatly needed data for improved und erstanding.

The diurnal cycle of outgoing longwave radiation OLR over the Sahara Desert is the strongest over the planet (Harrison et al., 1988) and has been studied by Duvel and Kandel (1985). Charney (19??) explained the flow over a desert in terms of the subsidence due to OLR. This explanation is verified by noting that the deserts stand out in maps of monthly-mean net radiation as having large negative values.

The deep convection regions of the Congo Basin and the Amazon also have very strong diurnal cycles. These are manifested by a pattern of clear mornings, with low clouds developing in the afternoon and developing into deep convection in late afternoon and evening, which clears by morning. The cycle operates as follows. Solar energy is absorbed at the surfce and much is converted into latent heat by vaporizing surface water and by evapo-transpiration. By afternoon, sufficient energy is stored at low level that when the sensible heat has developed an unstable temperature profile, the system erupts into a deep convection system in which the latent heat of vaporization lifts air from the boundary layer to near the tropopause. The system is a steam engine whichs absorbs solar energy, converts it to latent energy and then to potential energy. The lifted air then flows from the deep convective region at high altitude and is replaced by air flowing into the region in the boundary layer. Most of the water in the process is recycled several times before flowing out the Congo or Amazon River. Water which flows out the river is replaced by moisture in the incoming boundary layer flow. The air which has been lifted to high altitude then flows to the subgsidence regions of the globe. Thus, the coupled deep convection/ subsidence system is a heat engine with solar radiation in the deep convestive regions as the heat source and outgoing longwave radiation in the subsidence regions as the heat sink.

During boreal winter the deep convective regions move northward. The result is that in Africa the deep convection moves toward the Sahel, bringing the wet season. Meso-scale convective complexes MCC occur here frequently (Laing and Fritsch, 1997; Haile et al., 1994). MCC develop as the result of absorption of solar radiation to generate latent and sensible heat to form a set of deep convective cells. When this set of cells is organized by the synoptic scale pattern, the individual storms can coalesce into an MCC, which has a life of its own and lasts for several days. Such systems are not only of regional interest. From June through October, depressions moving with easterlies from West Africa across the Atlantic Ocean are carefully watched, as they can organize the high latent heat over the Caribean region into tropical storms and hurricanes.

Although not in the part of Earth which is viewed by GERB, the Great Plains of the United States are greatly affected by MCC (Mattox, 1980). MCCs were not discovered until geosynchronous satellite imagery made it possible to observe them, However, they are responsible for most summertime precipitation over the Plains. Furthermore, they develop into synoptic processes which then propagate to affect the eastward part of the US. For this reason, a GERB on a GOES would be extremely valuable for understanding these very important processes over the US.

Charney, J., 19??: Classical deserification paper.

Cheruy, F., R. S. Kandel and J. P. Duvel, 1991: Outgoing longwave radiation

and its diurnal variation from combined ERBE and Meteosat observations,

Part II: Using Meteosat data to determine the LW cycle, J. Geophys. Res.,

96, 22623-22630.

Duvel and R. S. Kandel, 1985; Regional-scale diurnal variations of outgoing

longwave radiation observed by Meteosat, J. Clim. Appl. Met., 24, 335-349.

Haile, M., J. R. Milford and G. Dugdale, 1994; Mesoscale Convective Complexes

in Sahelian Africa asa Observed by Meteosat, 255-266, Proc. 10-th Meteosat

Scientific User's Conference, Cascais, Portugal, 5-9 September.

Harrison, E. F., D. R. Brooks, P. Minnis, B. A. wielicki, W. F. Staylor, G.G.

Gibson, D. F. Young, F. M. Denn and the ERBE Science Team, 1988: First

estimates of the diurnal variation of longwave from the multiple-satellite

Earth radiation budget Experiment (ERBE), Bull. Amer. Met. Soc., 69,

1144-1151.

Laing, A. G. and J. M. Fritsch, 1997: The global population of mesoscale

convective complexes, Q. J. R. Meteorol. Soc., 123, 389-405.

Mattox, R. A., 1980: Mesoscale Convective Comlexes, Bull. Amer. Met. Soc.,

61, 1374-1387.

Minnis, P. and E. F. Harrison, 1984: Diurnal variability of regional cloud

and clear sky radiative parameters derived from GOES data, Part III:

November 1978 radiative parameters, J. Clim. Appl. Met., v. 23, pp

1032-1052.

Home