MODIS Data Processing Meeting: Hobart May 11-12 2004

Overview

This meeting was held to bring together Australian developers and implementers of MODIS base processing and product generation code with the goal of promoting cooperation, the development of standards and product consistency and minimising duplication of effort, with focus particularly on the Direct Broadcast environment.
 

Eight presentations were made and were followed in the afternoons with wide ranging technical discussions on subjects solicited from the floor. The discussions included the issues of archive management and ancillary data, and future acquisition systems. 

 

This report contains a terse summary of the presentations and discussions and links to the presentation files.

Attendees:

Alex Held	CSIRO Office of Space Science and Applications
Shanti Reddy	ACRES, Geoscience Australia
Peter Dyce	CSIRO Land & Water
Alan Marks	CSIRO Land & Water
John Woolner	ACRES, Geoscience Australia
Craig Smith	ACRES, Geoscience Australia
Anthony Rea	Bureau of Meteorology
Jean Rivard	Australian National University
Brendon McAtee	Department of Land Information,WA
Gary Quinn	ES&S
Ken Suber	CSIRO Marine
Edward King	CSIRO Earth Observation Centre
Ian Grant	Bureau of Meteorology
Peter Turner	CSIRO Marine
David Griersmith	Bureau of Meteorology
Ian Barton	CSIRO Marine
Glen Smith	CSIRO Marine
Neal Young	Australian Antarctic Division
Chris Rathbone	CSIRO Marine

Tuesday am – Presentations

Alex Held – Modis Applications (Presentation)

Overview of Modis and its context in Satellite Environmental Time Series

Overview of much EOC-related MODIS activity within CSIRO

Shanti Reddy – MODIS Processing at GA (Presentation)

ACRES have a raw MODIS data archive on tape. PDS data is transferred to BoM.  Data also goes to DLI.

Typical downloads from GA are around 80GB/month.

MS2GT errors have been fixed in the last year – related to an approx. 750m error in northing. 

NRT is possible with data from Alice Springs available in Canberra within about 2 hours of acquisition.  More difficult with Hobart on a slower link – 2-8 hours.

$150 for retrieval of a full pass – planning to go NRT/fully automated on this in which case cost will come down further.

L2+ products under development as part of the CRC-SI, with DLI

Ian Grant – MODIS activity at Bureau of Meteorology (Presentation)

Much work with Sun Photometers and Radiosondes to validate AOT and Total Water Vapour products

 

L2+ Land Products Issues

1. MODIS products have different aerosol algorithms – MOD04 (Aerosol), MOD09 (Surf. Ref) and MOD18 water leaving radiance
2. atmoco options – massaged inst. Code?
3. BRDF modelling data management

 

Additional stations are not going to be feasible for at least a couple of years – 2006 at the earliest.  Looking at consortia approaches.  Casey, Melbourne, Darwin.

 

Brendon McAtee – Progress in MOD09 Development (Presentation)

Much detail on porting issues for MOD09 code.

Require validation of MODIS-derived TWV & Ozone.

Ken Suber – MODIS Ocean Processing (Presentation)

Discussion of process of porting Ocean code at CMR.  Lots of ancillary data relevant potentially to land processing.

Peter Dyce – Sentinel

Brief description of system (no presentation).  Discussion of BoM taking over operational system and CSIRO maintaining a development system.

Tuesday pm - Technical Discussion

Ian Grant – BRDF Data Handling

1. Issues to do with re-gridding
2. Important to recognise that there are distinct steps in the DAAC process and the modularity should be retained in any DB code so that incremental improvements can be grafted in more easily.

Edward King – Ancillary Data Acquisition

1. Attempt to list all sorts of ancillary data either in use or potentially usable.
2. Edward to resolve as many of the unknowns as possible and then to report back to group via Peter Turner.
3. Possibilities for asembling a common collection to be investigated subject to opportunities identified.
4. Suggestion that RT ancillary data acquisitions be archived from now on, so the situation for time series doesn’t get any worse.
5. Suggestion that maybe advantages to be had in downloading once for Aust. – but then it was pointed out that costs of transfer between institutions erode any advantage.

McAtee – Validation Activities

DLI are really interested to find out what validation data sets exist – opportunities with both BoM historical records and some CSIRO-led EO-1 related field campaigns are possibilities.

Held – BoM Cooperation for generation of operational products

Arose in the context of BoM having the serious operational infrastructure required to deliver services like Sentinel robustly.  General agreement from CSIRO that such a mechanism would be desirable.  A list of products that the BoM would like to operationalise includes: WV, O3, Aerosol, dust, smoke, SST, clouds, sea ice, fog and low cloud, volcanic ash, polar winds, fire hotspots, grassland curing index.

 

Wednesday am – Presentations

Edward King – EOC MODIS Activities (Presentation)

EOC has significant terra land holdings and the CMR-obtained/facilitated collection of whole-of-mission MOD021km (L1B) for both Aqua and Terra.

Aim is to get the data into an on-line system

Gary Quinn – Satrax700 receiving system from ES&S (Presentation)

Description of ES&S X-band reception station, design rationale and realisation.  Overview of processing systems that goes on the backend – virtues of Python were extolled.  Clients are really interested in thermal anomalies (fire), SST and ocean colour, and aerosol (pollution). 

Brendon McAtee (for Stefan Maier) – Land Processing at DLI (Presentation)

Description of BRDF and atmospheric correction used in land product generation at DLI.  Some interesting details on pseudo-probabilities for change detection. Currently 2-3hours after the aquisition the near real time fire scar maps are available from MODIS/Terra (Aqua must wait for the definitive ephemeris). SMAC processing takes about 40-50 minutes for the land pixels of a typical overpass.

Wednesday pm - Technical Discussion

Shanti Reddy – Future Acquisition Systems

Alex – Vision for a National X-band network – Need something in QLD and NT.

Quinn - $300K 2-3m X-band antenna -> L1B product

Leasing or Hire/Purchase are feasible – includes maintenance.

ESS to prepare a proposal/straw man idea set for use by proselytisers.

Rationale – conflict resolution, backup, long term focus, critical mass, standardisation and consistency, customisation to client needs, ease of use, real time applications.

Common Software For MODIS

Ocean Products – SeaDAS likely to be preferred over NASA Oceans code in the future, although support for terra , while reputed to be forthcoming, is not here yet.

Land Products….

1. Need to compile list of s/w and specs on web site
2. Software developers to provide info to EOC for Modis Community WWW site (+ post message on Ausmodis mailing list).

Generally it appears that the coordination at earlier Australian Community meetings, and ongoing liaison with the international DB community is bearing fruit in that there is little duplication of effort and people are trying to make things work together in a coherent fashion.

Low Level Archives

DLI and GA archive Raw data (CCSDS Packets)

BoM archives PDS

EOC L1B (product)

Is this a bad thing?  Should the BoM archive Raw too – means that they need RS STPS before being able to use it.  However PDS is the most compact – decision was to continue that status quo. 

MODIS to NPOESS

Alex to follow up with DRO lab at GSFC to find out who is doing what, and how Australia can become involved with NPOESS/NPP ahead of time so that we are ready when it happens.  General assent that this is a good idea.

Next Meeting

There will be an applications-focussed meeting in Perth at the ARSPC conference in October this year.  Probable next meeting for this technically focussed group will be in Canberra in February 2005, hosted jointly by Geoscience Australia and CSIRO EOC.

 

Note: The next international DB meeting is believed to be in Beijing in the latter half of 2004 – need to keep an eye out on the DRO website at GSFC for notification of this.

Thanks to:

1. CSIRO Marine Research (Peter Turner) for organising and hosting the meeting.
2. CSIRO EOC for providing the lunches and tea/coffee
3. John Woolner and GA for the tour and inspection of the TERSS reception facility at Droughty Point and Castray Esplanade
4. CSIRO Office of Space Science & Applications (Alex Held) for support for the meeting dinner.
5. Participants and their Agencies for contributing so freely, informatively and cooperatively.