The Earth Observer

March/April 1996, Vol.8, No.2

NSIDC DAAC User Working Group Meeting Report

--Konrad Steffen, (koni@seaice.colorado.edu), Chair PoDAG, National Snow & Ice Data Center
--Ron Weaver, (weaver@kryos.colorado.edu), DAAC Manager, National Snow & Ice Data Center

BACKGROUND

The Polar DAAC User Working Group (PoDAG) met April 1-2 at Goddard Space Flight Center. This was the tenth meeting of the User Working Group, and the second as the NSIDC DAAC-only user working group. (Previously PoDAG served as a joint advisory group to the Alaska SAR Facility and NSIDC DAACs.) We provide here a very brief description of materials presented and the action items arising from the PoDAG. Please contact Ron Weaver if you have any questions about material presented.

NSIDC DAAC Update

Roger Barry, NSIDC DAAC Scientist, provided an overview of activities during the past 12 months. Highlights of his presentation included the following:

Total user requests filled by NSIDC during 1995 remained roughly the same as 1994. However, over 60% of the requests came from new users. SSM/I-derived products remain the largest portion of NSIDC data sets accessed.
Several new products are nearing completion. These include a combined 1978-95 snow cover extent-sea ice extent product covering the Northern Hemisphere; updates to the Historical Arctic Rawindsonde Archive; a Greenland ERS-1 SAR mosaic on CD-ROM (in conjunction with GSFC scientists); and weekly snow cover extent, 1978-present, derived from passive microwave data.
NSIDC is working with Mark Anderson, University of Nebraska, on the development of a sea ice melt onset product derived from passive microwave data.
NSIDC DAAC will acquire SMMR and SSM/I historical orbital brightness temperature data for 1978-95 from the Marshall Space Flight Center DAAC, due to the closure of the MSFC DAAC. Several other data sets will be transferred to NSIDC, such as the SSM/I NESDIS sensor counts for F8, F10, F11, Fleet Numerical Meteorology and Oceanography Center (FNMOC) antenna temperatures, and the Wentz antenna temperatures for F8 and F10. Transition will occur during the remainder of 1996.
The Data Center staff continues its collaborations with several IDS and instrument teams, including POLES and CRYSYS IDS groups, MODIS, ASTER, AMSR, AIRS, MISR, and GLAS Teams.
Routine data set production continued for approved passive microwave and AVHRR products. The production staff transferred, processed, and archived over 10,000 polar AVHRR scenes. SSM/I gridded products (TBs, and Sea Ice Concentrations) have been produced through August 1995.
NSIDC has completed a 16 bit version of the HDF conversion tool for SSM/I products. This tool is available on the FTP site (sidads). NSIDC will suspend production of new data sets in HDF until the tool's availability is adequate for the user community. NSIDC will launch a user survey to query the user community about the preferred data format (HDF vs. flat file structure).

Polar Pathfinder Interactions

Jennifer Francis reported on the Polar Pathfinder group (AVHRR, TOVS, Passive Microwave, Radar Altimeter, Radarsat) which met in Boulder (March '96), to determine standard formats (time, grid, filename, organization, variable names). They agreed to process first the Aug. `87-'93 time period, and then the pre '87 time period. Further, they agreed to merge all their geophysical products in a data cube (common low-resolution data set) with 100 km grid resolution on a daily basis (~500 Mb/year). TOVS data for April 87 - Nov. 88 are already archived at NSIDC (water vapor at 5 levels).

The group also has written an article (Roger Barry, et al.) that summarizes the Polar Pathfinder projects. It will be published shortly in EOS, Transactions, American Geophysical Union.

Glacier Mapping and Mass Balance Assessment

The Interdisciplinary Sciences Investigation on climate, erosion, and tectonics, led by Brian Isacks at Cornell University, has evolved to focus on mountain glaciers and the climate system. Mountain glaciers are sensitive indicators of climate change, and with new developments in satellite glaciology, multi-temporal high-resolution monitoring of certain glaciers could provide a valuable data set for climate system modeling. The theoretical, satellite-based, glacier/climate data assimilation machine was presented. This regional climate model would use passive microwave, visible, and radar satellite data as data inquest to derive mass and heat fluxes on the glacier surface. The selection of 23 present and planned glacier monitoring sites was presented.

MODIS Polar Products

At the November, 1995 MODIS Science Team meeting, NSIDC reported on two possible approaches for producing Level 3 global and polar grids for the MODIS snow and ice products. The preliminary work shows that errors are minimized when producing the Level 3 product directly from the Level 2 data, instead of regridding the Level 3 product. However, the proposed ISCCP-derived grid for MODIS Level 3 products may not be suitable for polar applications. Based on the concern raised by users, including the polar community, the Science Working Group for AM Platform (SWAMP) adopted a new policy which states: no single gridding scheme can be imposed on the EOS AM teams for Level 3 products -- there is too much variation in the requirements of the group (e.g., polar products vs. cloud products).

AVHRR Survey of the Modeling Community

A small user survey of the modeling community was conducted by Greg Flato, AES-Canada, to determine the usefulness of the proposed AVHRR data set. Results suggest that a 5 km data set would be ideal for regional climate modeling (10 responses total). The GCM modeling community showed no interest in the 1 km AVHRR data set.

AVHRR 1 km Archive

Three years ago the AVHRR 1 km archive was begun at NSIDC. Since the inception, a total of 78 different users made data requests, and the data use has grown steadily. The cost to the DAAC of the archiving and distribution of the 1 km data set has been around $80 - 120 K/yr. Increased user numbers are anticipated with the availability of the navigation software on the network. The generation of polar-wide geophysical products is now funded by the NASA Pathfinder project. As currently planned, the AVHRR Pathfinder products will be produced at three scales: 25 km, 5 km, and 1.25 km. All products will be generated on grids identical to the Equal Area SSM/I Earth (EASE) polar grid.

NOAA Operational Sea Ice Products

Rob Grumbine of NOAA's NCEP reported that National Center for Environmental Prediction (NCEP) now produces an operational sea ice concentration product based on the SSM/I F-13 passive microwave data, using the NASA team algorithm with the W. Abdalati correction coefficients in polar stereographic projection on a 25 km grid. The updated product is available on the following FTP account : 140.90.192.85 in the subdirectory pub/ice.

SMMR Sea Ice Intercomparison

Comparison between Pathfinder and GSFC SMMR brightness temperatures showed differences of up to 5 K (18 GHz/H). Large differences in Tb were found resulting from bad scan lines in the Pathfinder data set. The ice concentrations along the marginal ice zone differ consistently, up to 20 %, for the two data sets. The GSFC data set was "hand cleaned," and it is suggested that the Pathfinder data set needs some additional data cleaning.

SSM/I Intercalibration

The SSM/I F8, F10, F11 (1988-1995) data have been cross-calibrated by Frank Wentz to an accuracy of 0.1 K. The regression coefficients to convert F10, F11, to F8 have been released. The cross-correlation was only done for orbits with a time difference of less than 30 min. A detailed report is available from Remote Sensing Systems, Technical Memo 010395.

AMSR Sea Ice Products

Konrad Steffen will chair a study committee to determine which of several sea ice concentration and extent algorithms will become the at-launch algorithm for the Advanced Microwave Scanning Radiometer (AMSR). The AMSR is proposed for flight on the PM-1 platform. This meeting will be held in July 1996.

Passive Microwave Ice Tracking

Barry Goodison (AES Canada) reported on the cross-correlation of ice features (image matching) to derive ice motion based on 85 GHz SSM/I passive microwave data. This work has been published by T. Agnew, AES/Canada. Goodison went on to describe the CRYSYS IDS Team goals. The original focus of CRYSYS was on hydrology; this has shifted to the study of climate variability of cryospheric variables over a range of scales (regional to global). The research group concentrates on the development and validation of local, regional and global models of climate/cryospheric processes and dynamics, and on improving understanding of the role of the cryosphere in the climate system.

Don Cavalieri (GSFC) also reported on the recent progress in ice motion mapping using wavelet analysis (in space domain) to derive ice motion vectors from SSM/I passive microwave data. Comparison with buoy data showed good agreement for ice speed and ice direction in the Beaufort Sea.

Federation and Recompetition of DAACs

Dixon Butler (NASA HQ) briefed PoDAG on the current status of the federation of EOSDIS and recompetition of DAACs. The NASA study team, headed by Butler, is drafting a new scenario which embodies the concerns of the NRC Board on Sustainable Development.

PoDAG members raised several issues to be considered by the study team:

What role does a user working group play in the new Federation?
How does NASA maintain continuity and consistency in a long-term data time series (such as NSIDC's passive microwave sea ice concentration data set) in the recompetition scenarios?
PoDAG members suggested that new federation members should provide expanded information services rather than replacing existing DAAC services.

John Dalton, the Deputy ESDIS Project Manager, also spoke with the PoDAG membership. Dalton relayed his current view of recompetition and federation. In general terms, he echoed Butler's comments of the previous day. PoDAG membership again raised the issue of data consistency and continuity in a recompetition environment.

RADARSAT

Robert Thomas (NASA HQ) informed attendees that there might be a charge of U.S. $600 per image for RADARSAT data outside the Alaska SAR Facility (ASF) station mask, including those scenes down-loaded from the tape recorder within the ASF mask. NASA is negotiating with the Canadian Space Agency to resolve this issue.