OVERSEAS VISIT REPORT

CEOS WGCV IVOS Sub Group Meeting
CEOS WGCV Meeting
ATSR Core Group Meeting
RAL Model Meeting

Norway/UK, April 1999

Ian Barton - CSIRO Marine

Itinerary:

CEOS WGCV IVOS Meeting

The Infrared and Visible Optical Sensors Sub-group of the CEOS Working Group on Calibration and Validation met at the Andoya Rocket Range, Andenes, Norway during April 13-14. The agenda for the meeting is attached, and the official minutes of the meeting will be available in due course. Further details, including the minutes of the last IVOS meeting can be found at the WGCV WWW site.

The meeting started with the usual introductions, agreeing the agenda, minutes of the last meeting, and a review of action items.

Nigel Fox from the National Physics Laboratory (UK) informed the meeting of progress with a Mutual Recognition Agreement between major National Standards Labs. This international cooperation would assist with future maintenance of standards and transfer standards. NPL had been active over the last year providing traceable calibration for the AATSR instrument, GERB (the geostationary earth radiation budget instrument to fly on MSG-2), and CHRIS. This last instrument is a compact high resolution imaging spectrometer built by UK industry and to fly on an ESA platform as a demonstration package. CHRIS operates at wavelengths between 450 and 1000 nm. NPL have also calibrated a SOLSPEC instrument which will fly on the Space Station in 2003. This instrument will measure solar fluxes at the top of the atmosphere. NPL has also provided calibration for several aircraft instruments.

Bob Saunders from NIST (USA Standards Lab) gave a presentation of the NIST calibration of the instruments due to fly in the NASA EOS program. They had also been involved in calibration of the SeaWiFS instrument. Several reports of NIST activities were tabled. Bob also informed the meeting that the NIST report on the Infrared Inter-comparison held at Uni of Miami in 1997 would be available soon. Ian Barton and Fred Prata both attended this inter-comparison and calibrated their infrared radiometers against the NIST standard.

Phil Teillet of CCRS Canada gave a presentation of his results obtained using the QASAR concept. The results for the vicarious calibration of different Vis/NIR instruments showed a large dispersion which suggested that the techniques still required some fine-tuning. Note that later presentations showed similar results, and this issue generated much concern and discussion in both the IVOS meeting and the subsequent WGCV meeting. Phil also mentioned the recent changes in calibration coefficients published by NOAA. These changes, of more than 10 per cent, had occurred without any comments (or warning) from NOAA. Ross Mitchell at DAR should be informed of these changes, to see if they are consistent with his AVHRR-Watch program.

Alan Belward - the WGCV Chair - reported on the recent CEOS Plenary meeting held in India. Of particular interest to IVOS was the acceptance of South Africa as a CEOS member, and their possible inclusion in WGCV and IVOS activities. The IGOS strategy was now more widely based and the pilot projects were now themes of activity. The UNEP WWW site has details.

Alan Belward also mentioned a meeting to be held in London during May 26-28 to address issues related to cal/val over land surfaces. The meeting is a joint CEOS/ISPRS meeting and would be hosted by UCL. Details would soon be available on the CEOS WGCV and UCL WWW sites. Fred Prata may be interested in attending this meeting.

Aime Meygret from CNES, France reported on CNES activities. The 3 SPOT satellites were all still operational and SPOT-5 was due for launch in December 2001. CNES has used 20 cal sites over North Africa for both SPOT and VEGETATION. Both these instruments have on-board calibration, but results presented by Aime showed consistent differences between vicarious and on-board calibration. At this stage these differences were not reconciled.

David Smith gave a presentation on the calibration of the new AATSR instrument. This was now integrated on the ENVISAT platform and should be launched by the end of 2000. Dave was also keeping a watching brief on the ATSR-2 calibration activities including those of Fred Prata, La Crau, and comparisons with POLDER and GOME. The UK also has a joint program with NOAA where ATSR and AVHRR calibrations using the Libyan Desert are compared. Presentations on the ATSR-2 calibrations showed the same discrepancies with the vicarious and on-board calibrations as shown by CNES. It was evident that these groups should get together to try and resolve the issue. Some form of workshop - perhaps hosted by CNES - would seem appropriate.

The state of infrared calibration and SST validation was briefly discussed. The state of infrared cal/val seemed to be in good shape with on-board calibrators working well and SST validation progressing well. The value of the Miami radiometer inter-comparison was noted and a follow-on exercise in around 12 months should be encouraged. There is still a dearth of good radiometers and high quality ground-based radiometric validation data.

The meeting also discussed at length a possible recommendation to CEOS Plenary stating that future satellite instruments should provide data that are traceable to SI units. A recommendation was formulated and discussed in the WGCV meeting.

The next meeting location and date was not formally decided. One possibility is to meet just prior to the next WGCV meeting which was likely to be in India during February 2000.

CEOS Working Group on Calibration and Validation Meeting (WGCV-15)

This meeting was hosted by the Norwegian Space Centre and held at the Andoya Rocket Range (ARR), Andenes, Norway during April 14-16. An agenda is attached and many of the documents and country reports (including the Australian report) are available on the WGCV WWW site.

The meeting was opened by the Director of the ARR and the ALOMAR facility, Dr Kolbjorn Adolfsen. ALOMAR is a lidar laboratory which is situated on a mountain just adjacent to the ARR. The meeting was also welcomed by Per Erik Skrovseth, the Director of the Norwegian Space Centre.

The meeting was honoured to be addressed by the current Chair of CEOS - Dr Tillman Mohr from EUMETSAT. This was the first time that the CEOS Chair had attended a WGCV meeting. Dr Mohr passed on the thanks of Plenary to WGCV for their past efforts in the fields of calibration and validation. He also provided information on the status of the IGOS program. Dr Mohr discussed the relationships between the space component of IGOS, World Weather Watch, the Global Atmosphere Watch (GAW) and GOOS, GCOS etc. He also discussed UNEP's Global Environment Monitoring System (GEMS). The GOFC, and GTOS programs are to be discussed at a June meeting of the SIT.

Subsequent discussion with Dr Mohr on the operation of Plenary and the interaction between Plenary and the CEOS Working Groups were most valuable. Dr Mohr encouraged more WGCV contact with the CEOS Secretariat rather than waiting for the annual CEOS Plenary meeting. This avenue of communication was carefully noted by the WGCV.

Following the address by Dr Mohr there was a report from each WGCV Sub-group on their activities since the last meeting.

Yes-Louis Desnos described the objectives of the SAR Cal sub-group and reported that their next meeting would be hosted by CNES in Toulouse during October 26-29. A meeting flyer was tabled and copies can be obtained from Ian Barton or at http://www.estec.esa.nl/CONFANNOUN/99b02. Yves-Louis also suggested that the following SAR cal meeting should be held outside Europe and North America. Japan was suggested and he would explore that possibility.

The second sub-group report was presented by Jim Shiue who reported for the passive microwave group. This group would meet in California during September 1999 and would mainly cover TRMM and NOAA-15 activities. Future programs would include ENVISAT, ADEOS-II, EOS-PM-1 and NPOESS. Details of a proposed TRMM follow-on mission were also tabled.

Chris Mutlow reported on the IVOS meeting held just before the WGCV meeting (see above).

Ian Dowman reported for the Terrain Mapping sub-group and confirmed the details of the Land Cal/val meeting to be held in London during May 26-28. See IVOS report for details.

The Chair of WGCV (Alan Belward) reported on the CEOS Plenary. All 5 recommendations from WGCV were accepted. For other details see the IVOS report above.

The IVOS recommendation on the problems related to differences in on-board and vicarious calibration of Vis/NIR sensors was devolved to setting up a workshop of interested parties, perhaps hosted by CNES.

There was considerable discussion on the traceability of satellite measurements to international standards. This obviously has financial implications for the space agencies, and eventually the topic was held over to the next meeting by which time an action plan by the standards agencies would be developed.

CEOS had been offered some free display space at the UNISPACE'99 conference. Colour photos, overheads etc. showing the advantages of good cal/val should be sent to Alan Belward. Input for the WGCV slide set should also be submitted.

A session on Norwegian activities included presentations from four keynote speakers which was followed by a tours of the Andoya Rocket Range and the ALOMAR lidar facility located on a nearby mountain top. Further details are available from your CEOS WGCV representative.

Yves-Louis Desnos requested information on coastal radars operated within Australia. These could be used for validation of wind/wave products derived from SAR data. Contact points have been sent to Yves-Louis.

Country reports that were not on the WWW were presented from India, ESA and NOAA.

India reported that the ISRC1-B satellite had recently been launched. ISP-4 will be launched during May. This latter satellite carries an ocean colour sensor with 360m resolution and has a 2-day repeat giving global coverage. An AO for scientific exploitation would be released in May.

ESA reported on current and future missions based on ERS series and ENVISAT.

NOAA reported on the revision of pre-launch calibration coefficients for the MSU sensors that have flown on the NOAA series of satellites from TIROS-N to NOAA-14. NOAA also commented on the recent abrupt changes in the calibration coefficients for the short-wave channels on AVHHR. Details of why these changes occurred would be placed on the WWW.

The next WGCV meeting would most likely be in India (Bangalore) during February 2000.

ATSR Core Group Meeting - RAL, UK - 19/04/99

The minutes of this meeting will be available in due course.

The meeting started with approval of the last minutes and a run through the action items, most of which had been completed.

ESA reported that they can now provide access to Near-real-time ATSR browse images via the WWW. Users will need to register via the WWW. The latest 2 to 3 days images are available. Address is http://192.111.33.173/ATSRNRT/.

RAL reported a revised ATSR WWW page which gives information on both ATSR-1 and ATSR-2 sensors as well as movie loops of global SST data.

The inter-comparison kit for ATSR was under development. The meeting decided to go, ahead with processing 3 months of data in 1997. The aim is to produce a data set with matching ATSR and ground truth (buoy) data.

There was some discussion on AATSR data availability and this would be discussed further at the upcoming PSP meeting in May. NERC and CSIRO were adopting a common approach to gain free access to AATSR data.

MSSl Meeting to discuss RAL Model - 20/04/99

Mullard Space Science Laboratory, Holmbury St. Mary, UK - 20 April 1999

PRESENT:

Albin Zavody - Zavody Enterprises Inc.

Andrew Birks - RAL

Andy Harris - UKMO Hadley Centre

Chris Merchant - MSSL

Ian Barton - CSIRO

AGENDA

MODEL COMPARISONS

Four versions of the model were compared -

IJB - Ian Barton's version from CSIRO
CJM - Chris Merchant's version
AMZ - Albin's latest version as installed at Uni of Miami
RAL - Albin's nRADGEN model - latest version at RAL.

(a) All versions of the model use an exponential interpolation of water vapour with height between levels except CJM which uses a linear interpolation with height. Chris reported that both these led to some discontinuities with the slope and that a better scheme would be to apply linear interpolation of relative humidity with height. All agreed that this should be introduced into all the models.

(b) All models included the revised tropospheric aerosol subroutine as distributed in November 1997. The hybrid scheme developed by Phil Watts (RAL) was used by both CJM and IJB models. AMZ and RAL models were currently just using "absorption = extinction".

(c) All models include the revised stratospheric aerosol subroutine as distributed in November 1997. All models include the hybrid scheme for stratospheric aerosols.

(d) All models use the Clough Version 2.2 continuum formulation. This was still felt to be inadequate and there were also some concerns expressed as to whether the formulation was correctly coded in the models. The latest formulation - version 3.0 (perhaps not yet released) should be obtained. Meanwhile the implementation of Version 2.2 should be checked. (Done by CJM, and all is OK as far as 2.2 is concerned.)

(e) All models include the use of "msf" data files derived from the HITRAN data base. These are generated by Albin for different species, and wave-number ranges. IJB can include msf files which are for ozone absorption only. For all other models the ozone absorption is included in with carbon dioxide and other mixed gases. Albin has now generated new super-msf files which cover wider spectral ranges and include ozone absorption with the other relevant species.

(f) A range of techniques for introducing surface emissivity was evident in the models. IJB included fixed values for the ATSR nadir views and a parameterisation taken from Phil Watts' paper was used for the forward views. This parameterisation included effects of surface wind speed but did not account for any effects introduced by variations in water vapour or atmospheric transmittance. CJM has developed a full implementation of Phil Watts emissivity using the reflection enhancement. A better method would use atmospheric transmittance instead of water vapour and this should be developed. AMZ and RAL both ingest fixed values of emissivity from a data file which gives values for each ATSR instrument. These values are for a range of air-masses and view angles.

(g) All models include the ingestion of filter functions and rad-to-temp LUTs for each channel. AMZ and RAL now generate the LUTs for each model run. For ATSR-2 AMZ and RAL may use a filter functions and LUTs which are truncated at the ends of the available msf spectral ranges. The new super-msf files should mean that the full pass-bands of the filters should now be used. CJM and IJB both used the published forms of the filter functions. AMZ and RAL take account of the means by which these filter profiles were obtained - whether by photon or power measurement. Evidently this was different for ATSR-1 and ATSR-2. Thus the function for (at least) the 3.7 micron channel for ATSR-2 has been divided by wavenumber to account for this difference. The non- linearities for the detectors is assumed to be accounted for in the SADIST processor. However it was not clear whether the suggested spectral shift of 30 cm-1 was included for the 3.7 micron channels in the SADIST processor.

DISCUSSION

The meeting agreed that the water vapour interpolation should be linear with pressure. This would overcome the "scalloping" effect produced by the current scheme which employs an exponential interpolation with height.

The latest tropospheric aerosol subroutine was considered to be adequate, but there was some discussion as to whether the hybrid scheme of Phil Watts should be included for both the tropospheric and stratospheric subroutines. Generally it was agreed that the scheme should be included for the stratosphere but the meeting was uncertain about the troposphere. As a first step we agreed to undertake a sensitivity study to determine the effect of implementing the hybrid scheme in both the troposphere and the stratosphere.

Ian Barton presented two case studies which indicated that there may still be some under-estimation of the continuum absorption. Although it was confirmed that the current models incorporate the Clough Version 2.2 correctly, there was still some doubts. Every effort should be made to obtain details of the latest Clough formulation (Version 3.0) and to install this in all the models in the same manner.

Albin had recently developed super-msf files to cover the entire spectral ranges of the ATSR filters. All models should use these files. These files include water vapour lines, carbon dioxide, ozone, and other mixed gases. There was some discussion on the merits of including the effect of CFCs in the model (was there an action item to cover this ???).

The specification of surface emissivity showed the greatest divergence between the models. It was agreed that the full implementation of the emissivity as specified in Phil Watts' paper should be implemented in all the models. This would include the effects of wind speed and atmospheric transmittance.

The discussion about filter functions and radiance-to temperature LUTs created considerable confusion with the meeting. The CJM model generated LUTS from the full filter function as defined by the ATSR project. AMZ and RAL both use a truncated filter profile and a LUT developed with this truncation. IJB was uncertain whether his model used the same filter and LUT as AMZ or whether the filter profile and LUT were generated in the same manner as for CJM. Some checking and comparisons are required here. It was also reported that AMZ and RAL both normalise the filter responses when the filter responses were measured with a photon detector. This is not done by CJM and IJB.

The problems associated with apparent non-linearities in the 3.7 micron channel were discussed. There was a suggestion that this was caused by a 20 to 30 micron shift in the response of the detector. It was not clear whether such a shift had been incorporated into the published filter functions and whether it had been also incorporated into the LUTs.

The meeting then recognised that the issues addressed in the last two paragraphs were also an issue in the SADIST processor. The meeting agreed that some clarification was required on what was actually done in the SADIST processing. This included possible implementation of a finite black body emissivity for the two on-board calibration sources, the treatment of non-linearities, and the means for deriving rad-to-temp LUTs. Both ATSR-1 and ATSR-2 should be considered in this activity.

NEXT MEETING

The meeting agreed to re-convene during the June Workshop in Frascati. Given that Albin's attendance is critical to the success of these meetings every effort should be made to ensure that Albin can attend the Frascati meeting.

ACTIONS

Meetings at RAL to discuss ATSR Issues.

Discussions with Andrew Birks regarding problems with the Zavody model. I agreed to send copies of early Zavody programs back to RAL.

Discussions with Tim Nightingale about the results from the Miami workshop last year. I will send my data files which shows some variations in the NIST BB standard.

Discussions with Nigel Houghton regarding access to future ATSR data including ABT data and those for dedicated validation campaigns.

Appendix A - IVOS Meeting Agenda

CEOS Working Group on Calibration and Validation

Infrared Visible and Optical Sensors (IVOS) Subgroup

10th Meeting - 12 and 13 April 1999

Andoya Rocket Range, Norway

Draft Agenda

12 April

1400 Welcome

1410 Introduction and Logistics

1420 Meeting Objectives, Agree Agenda

1430 Agree minutes of the last meeting, Review Action Items

1530 Reports from CEOS WGCV and Plenary

1600 Coffee

1630 Country/Agency Reports

1730 Close

13 April

0900 Country/Agency Reports (cont)

1030 Coffee

1100 Discussion

1130 First Results from a Generalized Approach to the Vicarious Calibration of Multiple Earth Observation Sensors Using Hyperspectral Data - Phil. Teillet

1200 SPOT4 and VEGETATION Calibration - Aimé Meygret

Lunch

1330 AATSR Calibration - Dave Smith

1400 Results from a Test Site Survey with ATSR-2 - Dave Smith

1430 Discussion

1500 The way forward with test sites - how do we take the work forward ?

1600 Higher Level Product Validation

1630 Discussion

1700 Recommendations and report to WGCV

1730 Close

Appendix — B WGCV Meeting Agenda

WGCV 15

ALOMAR facility, Andøya Rocket Range, Norway

14 to 16 April 1999

Wednesday, 14 April, 1999

Registration from 09.00am

1. Welcome from the Norwegian Space Centre (Per Erik Skrovseth) and the Andøya Rocket Range (Kolbjørn Adolfsen)

2. Keynote address: Direction to the WGCV (Tillmann Mohr, Chairman of CEOS)

3. WGCV Chair's report (Alan Belward)

3.1. Report from 12th Plenary

3.2. IGOS update and status report

3.3. Action items from WGCV 14

4. Reports from Subgroups

4.1. SAR Calibration (Yves-Louis Desnos)

4.2. Microwave Sensors (Elena Lobl)

4.3. Infrared and Visible Optical Sensors (Christopher Mutlow)

4.4. Terrain Mapping (Ian Dowman)

5 WGCV Action Plan Review

5.1. Communication and outreach (including newsletter)

5.2. Test Site Dossier

5.3. Education

5.4. Validation Pilot Projects

6. Seafood buffet

Thursday 15 April 1999

7. WGCV's response to the IGOS pilot projects

7.1. Planning for the WGCV ISPRS meeting on land surface parameter validation (linked to GOFC requests)

8. Special session on Norwegian space activities

8.1. Norwegian Space Centre (Per Erik Skrovseth),

8.2. The Nansen Center (Lasse Pettersson)

8.3. The Andøya Rocket Range (Kolbjørn Adolfsen),

8.4. The ALOMAR facility (Werner Eriksen)

8.5. SAR wind Cal/Val for ENVISAT (Harald Johnsen, NORUT

9. Special session on Atmospheric Chemistry

9.1. Lessons learned from GOME and plans for Envisat E. Attema

9.2. Cal/Val work of GOME using the ALOMAR facility (Georg Hansen, Norwegian Institute for Air Research)

10. Tour of the ALOMAR Facility

Friday 16 April 1999

11. Country Reports (not previously distributed by web)

12. Summary

12.1. Interaction with IGOS projects

12.2. Recommendations to Plenary

12.3 Review of action items

13. Date and place of next meeting

Disclaimer