Report on Overseas Visit

Ian Barton

UK / Italy October 1997

Itinerary:

15-16 October. Hobart - London

20 October. London - Oxford

25 October. Oxford - London

26 October. London - Ispra, Italy

30 October - 1 November. Italy - Hobart

 

MEETINGS:

 

AATSR Program Steering Panel (PSP), 17 October, DoE, London

The meeting papers, consisting of the following, are available on request -

Prior to the meeting contact was made with Elio Grohovaz at DIST and he forwarded some details of recent discussions between DIST and DoE. Currently there are no issues of concern.

The current status of the AATSR program was presented - the visible and infrared calibration activities in the test facility at RAL was due to begin before the end of October. Some concern was expressed that the EGSE supplied Aspect included no maintenance contract to cover hardware failures. This had recently caused a problem with the failure of a power supply. Some concern was expressed about the effect of polarisation in the visible channels. This problem had arisen during the testing of the FPA. Three days had been put aside in the test schedule to cover polarisation - this would allow characterisation in the nadir but not at other angles.

Peter Dubock mentioned that the launch of ENVISAT had been moved to theend of 1999 but that did not allow any delay in the AATSR delivery date which was still January 1998. The EM integration was going extremely well, and the FM integration had started with microwave radiometer.

Ian Barton presented details of Australian AATSR data requirements including near-real-time data for the Australian Bureau of Meteorology (copy attached as Appendix 1). This presentation was discussed at length and highlighted some comments in the draft ESA data policy. DoE agreed to contact ESA and request clarification on some of the data issues. Concern was also expressed that the BoM was a commercial entity and thus may have to pay for data. DoE would request ESA to identify CSIRO and NERC as Category 1 data users through their provision of part of the space segment.

Lists of recent Australian activities and future validation plans were also presented to the meeting (Appendices 2 and 3).

BNSC presented a report on the planned ENVISAT Ground station at West Freugh - this would be the IK PAC (Processing and Archiving Centre).

The minutes of the PSP Meeting will be available in due course.

A copy of the Draft ESA Data Policy is also available on request.

AATSR Science Advisory Group (SAG), 22 October, Leicester University

The meeting papers are available on request: these consist of -

The minutes of the meeting will be available in the near future.

A presentation on the status of the ATSR-1 and ATSR-2 programs was delivered - both instruments continue to operate well. ATSR-1 is turned on for three days each seventy days.

Peter Kierley from ESA presented some concerns about the effect of polarisation in the visible channels of AATSR. There would be an attempt during vacuum tank testing to characterise these effects as well as possible.

Some effort may be required post-launch to assess the full implications of polarisation effects.

The next meeting was set down for April 28.

ATSR Core Group Meeting, 23 October, Rutherford Appleton Laboratory

The meeting papers are available on request: these consist of -

The minutes of the meeting will be available in the near future.

The main topic of discussion was the re-processing activities for both ATSR-1 and ATSR-2. For ATSR-1 RAL were waiting on the delivery of the missing tapes from ESA. They were expected before the end of the year and reprocessing would begin early in the new year. The processing would include the production of ABT data files as well as the ASST. A Master Request File had been established and selected sites would be provided with mission long high-spatial resolution data. These sites include several of the validation sites within Australia.

The ATSR-2 reprocessing had begun in spite of the ABT module not yet being included. It would be implemented as soon as possible. Meanwhile data are being retrospectively reprocessed starting with Cycle 25 (35-day cycles). Cycles 25 and 24 had been completed. Data from the master request files would be available via anonymous ftp sites. Users are to be notified of this arrangement in the next few weeks. When the reprocessing was complete those cycles missing ABT products would again be reprocessed.

The ATSR reprocessing includes the population of the UK browse facility.

The algorithm development meeting to be held the following day was briefly discussed and Chris Merchant from MSSL gave a short presentation on the current state of algorithm development for ATSR-1 and ATSR-2.

The next meeting was set down for 24 February 1998.

ATSR Algorithm Development Meeting, 24 October, Rutherford Appleton Laboratory

This meeting was comprised of a few "experts" to assess the current status of the algorithm development for ATSR-1 and ATSR-2 data. Chris Merchant from MSSL described six and four channel ATSR-1 algorithms that provided consistent data for the months of October 1991 and April 1992.

These were taken to be representative of maximum Pinatubo aerosol content and a well-dissipated aerosol. The six and four channel algorithms also provided the same SST fields to better than 0.1 K.

There was also some discussion on the extra information to be included in the ABT product. It was suggested that this should contain some information of where the "centre-of-gravity" of the contributing pixels occurred.

Chris Merchant uses a set of 496 coastal radiosonde profiles for his analysis. Ian Barton undertook to send him a copy of his 885 maritime sonde data.

 

OTHER DISCUSSIONS:

Albin Zavody: Discussions on developments with the atmospheric transmission model. New coefficients for treatment of the water vapour continuum absorption have been added as well as a revised treatment of the tropospheric aerosol absorption. A revised model was obtained for use with GLI, MODIS and GMS instruments. Future developments with SST algorithms were discussed as well as other uses of the ATSR and AATSR data.

Nigel Houghton and John Wright: ATSR data were obtained to cover the missing ASST orbit products. Data requests were lodged for ungridded data on behalf of A. Prata as well as for the data to cover Bill Emery's FRANKLIN cruise late in November. The long-term data requests for ATSR-1 and ATSR-2 were confirmed and procedures established for obtaining the regular ATSR data as it was reprocessed. Nigel Houghton has undertaken to supply a list of outstanding data requests for Australian users.

Phil Watts: Some discussions on participating in a future cirrus cloud experiment to be held in tropical Australia with the UK Met. Office.

Also discussions on a possible visit to DAR by Phil Watts in the coming year.

 

CASOTS (Combined Action to Study the Ocean's Thermal Skin) Meeting - JRC, Ispra, Italy, 28-29 October.

This two-day meeting follows on from the first meeting held in Southampton during 1996. The meeting opened with a presentation from Dr Kristina Katsaros which was followed by six sessions each addressing a different aspect of the topic. The final report will be available on the WWW in due course (http://www.casots.soc.ac.uk). The final session on future research directions and actions was chaired by Ian Barton, and a report on the session was provided for inclusion in the final meeting outcomes. A copy is also attached as APPENDIX 6 (below).

One of the major outcomes was the planning for an infrared radiometer inter-calibration to be held early in 1998. This would most likely be at the University of Miami, and would be organised by Peter Minnett. The possibility of including an SST Validation workshop at the same time would be explored. This could be a meeting of those planning to contribute to SST validation for GLI (NASDA, Japan), MODIS (NASA, USA) and AATSR (ESA., Europe). Australia is involved in all three of these instrument validation campaigns.

APPENDIX 1.

Australian data requirements - Research

High spatial resolution data

For data product validation, process studies, algorithm development and research applications.

(a) Land sites: Hay, Amburla, Tinga Tingana, Broome, plus others as they are developed. Data from all AATSR passes required.

(b) Ocean sites: Townsville, Perth, Cape Grim, and others as they are developed. Data from all AATSR passes required.

(c) Shorter term sites to cover field experiments, ship cruises, special requests, etc. These data will be for both validation and research applications.

Averaged products

All averaged (ASST, ALST, ABT, etc) products globally and orbit by orbit.

Australian data requirements - Operational

These requirements are under development with the Australian Bureau of Meteorology (BoM) - but will be similar to those of the European meteorological services and the data may be obtained and used in a similar fashion. The BoM is expecting access to the data in the same manner and time-frame as the UK Meteorological Office and other European agencies.

APPENDIX 2.

Recent Australian activities

1. A series of ATSR-related meetings was held in Townsville during April. These included a mini-SAG, and details of these meetings are included in the AATSR PI's report.

2. During the Townsville meetings a validation excursion was held to the outer barrier reef. The results of this case study have been incorporated into a short paper submitted to "Marine and Freshwater Research".

3. Under the auspices of the CEOS WGCV, a one-week intensive campaign of ground and satellite data collection was held during July-August 1997. Details are available in a separate report.

4. The ACE-1 experiment - a joint collaboration between CSIRO and Nottingham University - was reported at the mini-SAG in Townsville.

5. A cruise on the RV FRANKLIN has just been completed. Infrared data were obtained with Fred Prata's revamped "SON-OF-AVADS" IR radiometer as well as a TASCO radiometer. Radiosonde data were also collected during satellite overpasses. The weather was not too kind and the number of coincidences with ATSR-2 is not yet known.

6. A new high-quality infrared radiometer is under construction that will be dedicated to ship measurements for satellite data product validation. This will be similar to ``SON-OF-AVADS".

APPENDIX 3.

Future Plans

1. Bill Emery is currently spending ten months sabbatical in Hobart. He will partake in a short cruise on FRANKLIN late in November. The cruise track is from Townsville to Tasmania. The new CSIRO radiometer will be tested as well as TASCOs and other radiometers. Data will be available for ATSR-2 data validation.

2. We are exploring the possibility of holding a second IR radiometer inter-comparison. This will be further advanced at the CASOTS-II meeting in Italy during late October. The previous inter-comparison at CASOTS-I did not include the "best" radiometers. At the next inter-comparison we hope to include -

A tentative time and location is March, 1998 in Miami.

APPENDIX 4.

Long-term validation plans for AATSR - SST

In conjunction with other Australian institutes CSIRO is planning a project that aims to develop techniques for the precise validation of SST as determined from data provided by environmental satellites. The project will develop a long term SST validation data set for assessing the accuracy of SST as measured by several instruments including AVHRR, ATSR-2, AATSR, MODIS and GLI.

The project is based on three separate activities -

(a) The Australian Institute of Marine Sciences (AIMS) is undertaking the deployment of instrumentation on a wave piercing ferry that operates a daily service between Townsville and the outer Great Barrier Reef. On the outward journey the ferry will be under way at 1030 local time - the overpass time of many of the targeted satellites. The instrumentation will include an infrared radiometer mounted on the ferry's bow and a measurement of bulk water temperature by sampling the engine intake water.

This location has minimal cloud cover during the southern winter and the coastal zone is also often cloud-free during the summer.

(b) The collection of validation data off the coast of Fremantle in W.A. This activity has been underway for some years now as part of CSIRO's research into the Leeuwin Current and in support of the fishing industry in W.A. Data collected are used to verify the SST obtained using AVHRR data. The current measurements include bulk temperature sampling at about 3-weekly intervals at a station off Rottnest Island and the collection of data on a new transect just off the coast near the Marmion Labs. It is planned to augment these measurements by installing instruments on a ferry which operates a daily service to Rottnest Is.

Initially a bulk temperature will be monitored, but funds are being sought to include a radiometric measurement in the same way as is done in Townsville. It may also be possible to increase the measurements made on the transect off Marmion.

(c) The development and deployment of instrumentation on research ships of opportunity. These will include FRANKLIN, SOUTHERN SURVEYOR and AURORA AUSTRALIS. The Antarctic CRC have a project which involves the collection of thermal IR data over the East Antarctic pack, using radiometers connected to the Aurora Australis. Such data will help validate surface physical temperature data derived from channels 4 and 5 of the AVHRR.

All of these activities will provide a long-term high-quality SST validation data set over a range of climates and seasons.

CSIRO continues to operate two ground based measurement sites over uniform inland areas. These sites are instrumented to continually measure the radiation balances at the earth's surface and are ideal for the vicarious calibration of satellite sensors and inter-comparison between different sensors.

A new site is planned for the Strezlecki Desert which provides an invariant site for the vicarious calibration of short-wave satellite radiometers. This will initially be used for the calibration of AVHRR and the visible channels on the geostationary GMS satellite.

CSIRO is also looking at establishing another CIGSN site in the tropics - perhaps near the north-west coast near Broome.

Details of these activities have been incorporated into the AATSR Validation Plan.

APPENDIX 5.

The CEOS IR Inter-comparison Project

 

At the WGCV meeting in Germany in December 1996 the WGCV established two pilot projects to demonstrate the value of frequent cross-calibrations between similar satellite sensors.
The two projects were selected to cover the two main electromagnetic bands used in earth observation - the visible-near infrared region and the thermal infrared. These projects are now well under way with the latter now in the data analysis phase.

The main objective in the infrared project was to collect an intensive set of ground-based data to compare with the different satellite data and to assist in the satellite data inter-comparison. A site in the Coral Sea near Townsville, Australia was chosen because the necessary infrastructure was available to support the ground campaign. A one-week period in July-August was selected to maximise the likelihood of cloud-less skies - the high cloud activity associated with the Australian monsoon is well north of Australia during the austral winter.

Two ships were used to collect the ground-based data; a fast tourist ferry which, on four days, travelled from Townsville to the outer Great Barrier Reef and carried an automatic infrared radiometer, and a smaller catamaran which operated between Townsville and 20 km north of Magnetic Island on nine occasions during the week. The smaller catamaran collected data using three infrared radiometers, two in-situ thermometers for bulk sea surface temperature and a third thermometer for air temperature. During each of the nine excursions the Bureau of Meteorology launched a radiosonde from the Townsville Airport to measure vertical profiles of temperature and humidity in the atmosphere above Townsville. These data will ultimately be used in a transmission model of the atmosphere to relate the surface and satellite measurements. Meteorological stations situated at three locations on the coast and Great Barrier Reef supplied additional meteorological data including wind speed and direction, surface pressure and air and sea temperatures.

The radiometric and thermometric measurements of the water temperature are both required because in normal conditions the top layer of the ocean (the "skin layer" which is less than a millimetre deep) is cooler than the bulk temperature below. Satellites with infrared radiometers actually measure the skin temperature while traditional SST measurement methods provide the bulk temperature.

During the one-week period regular data were collected from the Japanese Geostationary Meteorological Satellite (GMS) situated above the equator at a longitude of 140 degrees E. These data consist of hourly images from both the visible and infrared channels of the radiometer on the GMS satellite. All data from the two NOAA polar-orbiting meteorological satellites were also collected. Again these include both visible and infrared images from the AVHRR sensors on the NOAA-12 and -14 satellites. Data from the Along Track Scanning Radiometers (ATSR) on the ERS-1 and ERS-2 satellites were also recorded during the week. These data were transmitted to ground stations in Europe and have subsequently been transferred to Australia via electronic data transmission.

A preliminary analysis has been carried out on the data with the following tentative results:

1. The skin-bulk temperature difference was usually around 0.2 to 0.3 degrees. A more detailed analysis may provide some information on which parameters control the magnitude of this difference.

2. The infrared radiometers used on the catamaran (two TASCO radiometers and one Everest 4000A) showed good agreement with each other.

3. The AVHRR data provided a good estimate of the bulk surface temperature. The accuracy appears to be better than 0.5 degrees. There is some evidence that the estimates from NOAA-14 data are about 0.8 degrees warmer than those from the NOAA-12 data.

4. The GMS data indicate that the split-window brightness temperatures at 11 and 12 microns are too close to each other. This causes an under-estimation of the derived SST at times when there is considerable water vapour present in the atmosphere.

5. The ATSR data show good agreement with the ship measurements of the radiometric surface temperature. Accuracies were generally within 0.3 degrees.

 

Radiosonde data and radiometer data from the tourist ferry are yet to be analysed - they will extend the coverage of the analysis to well beyond the coast line.

Data analysis will continue with detailed comparisons between the ship and satellite data. Brightness temperatures measured by the satellite as well as the derived SST values will be considered.

Further satellite data sets are expected from the Chinese and Indian geostationary meteorological satellites as well as the DMSP and LANDSAT satellites of the USA.

APPENDIX 6.

Draft report on the last session at the CASOTS-II workshop.

Report on Final Session at the CASOTS Meeting

The meeting agenda flagged the following issues to be addressed in this final session.

 

Scope:

1. Bibliography

2. Data catalogue

3. Outstanding questions about the skin-bulk temperature difference (DeltaT)

4. Directions of future research effort

5. Communications.

 

Questions:

1. Gaps in knowledge of DeltaT

2. Identify the processes which control DeltaT

3. Are collaborative programs required?

4. Suggestions for a review paper or a journal "Special Issue"

5. Need for a DeltaT data-base

 

Outcomes:

1. Future research programs

2. A communications strategy

 

Bibliography: Ian Robinson reported that the CASOTS bibliography was taking shape and would soon be placed on the WWW. He encouraged all participants to ensure that a list of suitable references was forwarded to Duncan Baldwin or himself. It would be preferable if those sending references undertook some preliminary editing of their lists to ensure that only relevant references were included and that a simple format was used. This would minimise the extra work required to collate the final list.

Data Catalogue: Ian Robinson described the status of the data catalogue and again asked for input from the general community. There was some discussion about the content of the data catalogue but it was agreed that meta-data describing the contents of each entry were required. It would also be necessary to provide a "point-of-contact" for each entry to assist those wishing to use the data catalogue.

The session leader then presented information describing the anticipated scope of the discussions in the session. The plan was to break the session into seven areas identified below, and a brief introduction to stimulate discussion in each of the topics was presented.

 

1. The Skin Effect - Small-scales: This topic would cover the use of high spatial resolution instruments (IR cameras) in studying the effect of breaking waves and surface skin renewal. It would also include those small scale theoretical studies attempting to model the physical processes at the air-sea interface. It was not clear that the relative effects of surface emissivity variations, foam effects, surfactants, and breaking waves were well understood. A full understanding of these effects would be required before a reliable parametrisation between skin and bulk temperatures could be formulated. These effects would also need to be scaled up to satellite pixel (1 Km) and model grid (100 Km) scales if we are expecting numerical models to differentiate between skin and bulk temperatures.

Discussion: The main discussion centred around the requirements and perceived needs of numerical modellers. Some form of parametrisation would be needed if skin temperature was to replace bulk temperature as an input to the models. The assimilation of the data into models was also recognised as a major problem. It would be important to cross-link work in this area to other activities related to ingesting satellite data into models. It was agreed that the modelling of small scale processes was crucial to our understanding of DeltaT. It was also important to relate the air-sea interface to the bulk temperature below. Data assimilation into all forms of models is required - ocean, atmosphere and coupled for both forecasting and climate studies.

2. The Skin-Bulk Relation - Larger Scales: At these larger scales investigations would be more of an empirical nature. High quality data under many different environmental conditions will be required if an empirical relation for DeltaT is to be established. It was evident that the current formulations are still inaccurate in many situations. The possibility of adopting an incremental approach was identified with a progression of studies going from "laboratory" to "pier" to "platform" to "ship". In many cases investigators start out on the most difficult of platforms (ships) before instruments and simple theories have been tested.

Discussion: It will be necessary to forge links between those working at the small scale (e.g. using IR cameras) and those dealing with wide field-of-view radiometers in which some time and spatial averaging is used. It is these latter measurements which relate to the skin-bulk temperature difference as determined by satellite radiometers. The importance of DeltaT in other studies was also flagged. These include gas exchanges across the air-sea interface with carbon dioxide being the most important. DeltaT could also be modified by surface slicks and these need to be better understood.


3. Large-Scale Data Analyses: The possibility of using existing data sets in large scale analyses of DeltaT was proposed. Data sets from AVHRR, ATSR, SSMI, NSCATT, SCATT, and other satellites were now available and could be used to investigate relations between skin and bulk temperatures and the different SST products derived using theoretical and regression algorithms. Improved atmospheric modelling now means that good theoretical algorithms can be developed and these can be used to compare different satellite-derived SST values. Using the data sets it could be possible to compare DeltaT dependence on wind speed, solar insolation, time of day, etc.

Discussion: Bill Emery noted that such a study was already under way using AVHRR data and an SST algorithm developed by Peter Schluessel. Wind and water vapour could be obtained from SSMI data. Further analyses using existing data sets could provide valuable information on DeltaT parametrisation. The possible use of wind (and other) fields from NWP and ECMWF analyses was identified.

4. Diurnal Thermocline (Rain Layers): It was noted that since the first CASOTS workshop there has been an improvement in the understanding of the importance of these phenomena in DeltaT studies. The skin-bulk temperature difference is confounded by these oceanic surface features and it is important to specify a depth when referring to DeltaT - especially under light wind conditions and after rainfall.

Discussion: The main discussion on this topic was the importance of recognising the existence of these phenomena. If skin temperature is to be assimilated into numerical models then the models should also be able to recognise the existence of these features. Model wind fields, cloudiness, and insolation are obviously required parameters. It may also be necessary to model the complete mixed layer as well as the near-surface diurnal thermocline layer.

5. Data Collection: More high quality data bases are required to enable the testing of new parametrisations. Investigators should explore the opportunities provided by ships of opportunity - these platforms are able to provide regular data over large areas compared to the poor coverage of short dedicated campaigns. In all cases it is important to provide detailed metadata as well as highly accurate data using reliable calibration techniques.

Discussion: It was agreed that more data were required but that we should also ensure that data were collected over different geographical areas, climates and conditions. Some ships of opportunity (e.g. ferries) would only provide data over the same transect throughout the year, and this may not be sufficient as some locations may introduce biases through anomalous local conditions. The lack of a reliable towed thermistor was identified. Individuals had developed different techniques (seasnake, tin-hat, and others were mentioned) but all suffered from "wave-skipping" at normal cruising speeds and wave conditions. Richard Meade (?) had developed a system that can be towed outside the ship's wake and had been tested off the east coast of the USA. This system may be worth further investigation. Andy Jessup mentioned that FLIP was an excellent platform for undertaking DeltaT studies. FLIP was stable enough to be used for eddy flux studies - the platform also provided many of the basic measurements required to support such investigations. The importance of water vapour, aerosols and latitude (location) in relating satellite and ground based measurements was stressed. Aircraft were also mentioned as being useful platforms for radiometers as they could easily search for cloud-free areas as well as supplying "top-of-atmosphere radiances for comparison with satellite data. Aircraft radiometers could also, be used in the validation of atmospheric transmission models. It was suggested that details of instruments and platforms could find a useful place on the CASOTS data base.


6. Future Satellite Data: The future assimilation of skin temperature data into models requires two types of space platforms -

(a) orbiting satellites for precise absolute skin SST (ATSR, MODIS etc.), and

(b) geostationary satellites to follow diurnal variations, i.e. to accurately

identify the existence of diurnal thermoclines. The geostationary instruments should include split-window channels and 12-bit data transmission. These are stringent requirements and some degradation in spatial resolution may be required to obtain sufficient accuracy from geostationary orbits.

Discussion: Satellite instruments into the next decade have all been approved and such demands may not be acceptable to the instrument providers. GMS has split-window channels with 8-bit accuracy while the GOES satellites have 10-bits. The details of MSG were not known. It was agreed that some information on the diurnal variation of skin temperature would be required if these data were to be assimilated into climate and forecast models.

Future satellites could also carry higher spatial resolution imaging radiometers which could assist in cloud detection.

7. Future Meetings: Possible meetings were identified and discussed.

(a) Peter Minnett, Tim Nightingale and Ian Barton were planning an inter-comparison of their infrared radiometers early in 1998. This could be seen as a follow-on activity from that carried out at the CASOTS-I workshop. Instruments to be compared would be SISTeR, M-AERI, the CSIRO radiometer, TASCOs, Everests and any others that might be available. The first three instruments were not included in the CASOTS-I comparison. A tentative location and time would be Miami in February 1998.

(b) Ian Barton proposed an SST validation workshop to coincide with the Miami inter-comparison. Those associated with SST validation for GLI, MODIS and AATSR could meet to discuss sharing of resources and data, the development of common techniques and a discussion centred around the instruments' validation plans (all of which exist in some form).

(c) Ryuzo Yokoyama flagged the possibility of an SST meeting in Japan during June 1998. This would be centred around the results of the MUBEX campaigns but would also focus on general validation and SST derivation issues. It would also look at the use of OCTS data for SST retrieval.

(d) Other meetings had been flagged by several attendees during the workshop -

(i) PORSEC in Japan during July 1998 (details from Bill Emery and David Llewellyn-Jones)

(ii) IGARSS in USA during July 1998 (details from ??)

(iii) COSPAR in Japan during July 1998 (details from Nagaraja Rao, NOAA)

(iv) AMS in Paris during May 1998 (Andy Harris has some details)

(e) The possible need for the CASOTS community to meet again in about 2 years time was put forward - but no firm commitment or suggestions were forthcoming. The existing CASOTS partners would meet after the workshop to discuss the way forward.

Disclaimer