CLIMAP

CLIMAP Activity Overview

The CLIMAP project is a pre-operational project to pave the way for operational use of atmospheric sounding data in meteorological weather forecasting and climate monitoring. CLIMAP is funded by the Commission of the European Communities under the Environmental and Climate Programme, with TERMA as coordinating partner, partners being the meteorological offices in Denmark (DMI), The Netherlands (KNMI) and The United Kingdom (Met Office), and IEEC from Spain. The project was initiated in January 1998 and finished in June 2000.

Figure 1: CLIMAP System Overview

During the project TERMA and DMI have developed a pre-operational data processing/data distribution system to process and distribute GPS occultation based atmospheric measurements (bending angle, temperature and water vapour pressure profiles) from the Danish ØRSTED satellite. Ground based GPS measurement supports derivation of integrated atmospheric humidity as processed by an IEEC provided chain. ATOVS/TOVS vertical profiles are processed by KNMI. The products can be distributed/made available to accredited user groups associated with the Global Climate Observation System (GCOS) and World Climate Research Programme (WCRP). Once received by the meteorological offices, the sounding data shall be assimilated into climate and numeric weather prediction models. The Met Office and DMI have established data assimilation systems.

The pre-operational system has been validated and recommendations made for the efficient deployment of an operational Radio Occultation processing system.

Finally the project has developed a concept for a fully operational system to provide services to the climate and environment monitoring communities at large, including the European Environmental Agency's Topic Centre for Air Quality.

The proposed concept for processing the data received from the various space- and ground based-instruments is outlined in the figure above. It illustrates the overall approach of first processing the satellite data to produce a new set of data which may subsequently be assimilated into meteorological and climate models. Results from these observational data may then be distributed to the international community.

Workshops have been arranged throughout the project to inform the user community of the project status and provide for discussions as a basis for changes and enhancements.

Selected RO based data sets have been validated by comparison with analysed and short-period forecast fields data from the Met Office and ECMWF NWP systems: No GPS data have of cause been assimilated into these comparison fields but (A)TOVS will have been.

(A)TOVS based data sets are compared to forecast data from HIRLAM NWP model and to GPS based RO data sets.

The table 1 below further details the data acquisition, data processing, validation and the data assimilation activities and responsibilities.

Major administrative-technical responsibilities for the CLIMAP project have been the following

TERMA executed the overall project coordination and reporting. This included establishing a WWW access to the project description, to the major documentation and to a project database with selected data formatted according to the agreed and documented formats.
KNMI coordinated workshops.
The Met Office coordinated the definition of data exchange formats for assimilation

Source	Processing chain	Processing chain Output and data validation	Assimilation with model for Numerical Weather Prediction (NWP)
LEO satellites with GPS receiver measuring phase delay ('ray bending') due to atmosphere as a function of signal ('ray') height above earth. LEO satellites used: - GPS/Met, old data not operational - ØRSTED, with TurboRogue receiver	A Radio Occultation (RO) Processing Chain using satellite recorded data to be established by DMI & TERMA: Based on on-board receiver GPS occultation and tracking data plus simultaneous fiducial ground station GPS tracking data. The aim is to produce a semi-operational chain demonstrating the on-line delivery of data.	'Vertical' profiles of 'bending angles' (level 1b product) or derived level 2 products: Refractivity, Temperature and Water Vapour Pressure profiles. Data validation is performed by comparison as follows: DMI to ECWMF profiles generated using other data Met Office to profiles generated by own NWP model using other non-GPS data	DMI established assimilation by means of Optimal Interpolation Analysis using level 2 products. The model is HIRLAM and the Northern Atlantic region is considered. Met Office established assimilation based on 1dVAR input of level 1 products. The model is by Met Office and it is global.
LEO satellite with infrared and microwave wide-band instruments measuring atmospheric radiance with frequency ( HIRS, MSU or SSU ) LEO satellites used: - NOAA, early instruments, old data - NOAA, advanced (A) instruments	TOVS and ATOVS Processing Chains to be established by KNMI: Processing chain 1: Starts with TOVS radiances from NOAA-NESDIS followed by a 1DVAR-retrieval leading to vertical temperature and humidity profiles in the HIRLAM area. Processing chain 2: Starts with raw (A)TOVS radiances received locally at KNMI, also followed by a 1DVAR retrieval leading to temperature and humidity profiles. The last part of the processing chain is common and consists of encoding of the resulting profiles in BUFR-format and storage in the KNMI archive. The aim is to produce a semi-operational chain demonstrating the on-line acquisition, processing and delivery of data.	Temperature and humidity values on the 40 pressure. The profiles will be chosen so that they can be collocated in time and space with the Ørsted profiles. Validation of the temperature profiles obtained by 1DVAR is done in the first instance by KNMI comparison with HIRLAM analysis and forecast fields.	DMI established assimilation by means of Optimal Interpolation Analysis using level 2 products. The model is HIRLAM and the Northern Atlantic region is considered.
GPS tracking data are retrieved from ground based stations.	A processing chain using ground received GPS data to be established by IEEC for generating the integrated vertical atmospheric phase delay. The aim is to produce a semi-operational chain demonstrating the on-line acquisition, processing and delivery of data.	The phase delay is the basis for deriving the Integrated Profile of Water Vapour (IPWV) contents. Data are validated by IEEC through comparison to profiles established by radar measurements, etc. The Met. Office will also validate the ground-based water vapour (actually TZD) measurements provided by IEEC against NWP model fields.	The IPWV is use as input in 3-4 d VAR assimilation in preparation (not part of CLIMAP). Total Zenith Delay is developed as input to The Met. Office's operational 3DVAR mesoscale assimilation (not part of original CLIMAP proposal).

Table 1: Activities and responsibilities of CLIMAP

The table 2 below provides an overview of the final status of activities.

Company	Planned activity description	Status at end of CLIMAP-project
TERMA & DMI	RO satellite Data Processing, including demonstration of on-line data acquisition, processing and delivery. Validation of output.	ØRSTED data delivered with about 50 out of 250 possible occultations a day. Due to the antenna size, receiver and possibly interference problems the data are not of operational quality or sufficient, e. g. L2 usually missing or of bad quality, L1 often stops above the Tropopause (60 %). Processing chain is performing the level 0, 1a and 1b processing. Due to quality of ØRSTED data the 1b residual phase corrections are constrained to rough POD and ionosphere corrections with artificial L2. Chain not automated. Difference between using artificial L2 and real L2 is validated by TERMA and DMI Considering validation results using artificial L2 for Ørsted Considering comparing GPS/Met based artificial L2 to real GPS/Met based L2 The use of artificial L2 seems to provide good results but the conclusion is as yet TBD. Data from period Feb. 3-22 processed and validated by DMI comparing temperature profiles to ECMWF profiles based on non-satellite data. Results of the validation is a maximum deviation of 5 ºK and best 2ºK up to 30 hPa (25 km) where the non-satellite data stops. Above 40 km model influence makes occultation data non-useful. As for GPS/met much more data are needed for assimilation (two high gain antennas needed on satellites for a start). The same data was validated by Met Office through comparison to Met Office model generated data, refer below. Profiles (Bending angle, temperature) from February 2000 available via the Web/Internet from the CLIMAP project database. Raw old GPS/Met data very scarcely available. Attempt to revive GPS/Met as a reserve stranded for unknown reasons, probably due to economy and satellite exploitation politics.
KNMI	TOVS and ATOVS Processing Chains to be established. The aim is to produce a semi-operational chain demonstrating the on-line acquisition, processing and delivery of data. Preliminary tests on the obtained 1DVAR processed temperature and humidity profiles have been performed.	Off -line TOVS and ATOVS Processing chains available but not completely automated. At a certain moment a choice was made to focus on archived TOVS data instead of locally received ATOVS data. 1DVAR-processed TOVS data for some dates in Feb. 2000 will be available in order to do collocations with the Ørsted data from the same period. This is based on TOVS radiances from NOAA-NESDIS, obtained from the ECWMF MARS database. Data validation performed by KNMI through comparison with HIRLAM model fields has been done for some cases. Comparisons with some Ørsted profles in Feb. 2000 are performed. Results of validation are TBD. Almost all of the processing and interface software is ready and can be obtained by all users that are interested.
IEEC	A processing chain using ground received GPS data to be established by IEEC for generating the integrated vertical atmospheric phase delay and Integrated Water Vapour. Data will be validated by comparison to radar measurements. The aim is to produce a semi-operational chain demonstrating the on-line acquisition, processing and delivery of data.	Chain established but not yet fully automated concerning data acquisition and processing. Data output validated by IEEC comparison to delay established by radar measurements. Result of validation is TBD (? Available at IEEC Web page on GNSS based WV)
MET OFFICE	Preparation for an execution of assimilation of RO data into MET OFFICE global model currently using field data and other satellite data. Validation of RO data input. Validation of ground reception based GPS data	Performed occultation input analysis using old processed GPS/Met data. Demonstrated usefulness of these data in the northern hemisphere: Standard deviation within 1.5 º K from profiles of other data generated by model. No model execution with new radio-occultation data due to insufficient quantity of data. Performed validation of RO product through provision of and comparison to atmospheric temperature profiles for Feb. 2000 using own global NWP model. Results are TBD. Additionally analysed impact of spherical asymmetry error using radio-occultation data processed assuming symmetry. The Met. Office performed validation of ground-based GPS product though provision of and comparison with equivalent data derived from own mesoscale NWP model. Results TBD.
DMI	Preparation for and execution of the assimilation of RO and TOVS/ATOVS data into DMI NWP model (HIRLAM) for northern Europe.	Validated assimilation with old processed GPS/Met data. The period was only 14 days and the impact was neutral. TOVS data assimilation not performed within CLIMAP. By end of project not sufficient data available.

Table 2: Final Status of CLIMAP Project