--E. Lobl (elena.lobl@msfc.nasa.gov), AMSR-E Science Team Coordinator,
Earth System Science Laboratory, University of Alabama in Huntsville;
EOS PM-1 AMSR-E homepage: wwwghcc.msfc.nasa.gov/AMSR
An AMSR-E Science Team meeting was held on July 7, 1998 at the Sheraton Hotel, Seattle, WA, in parallel with IGARSS. The main purpose of the meeting was to give our team members a chance to present their latest research results. Besides usual team business, there was a presentation on ice sheets by D. Winebrenner, Applied Physics Laboratory (APL), University of Washington, Seattle, WA, and on the status of the Polarimetric Scanning Radiometer (PSR) by A. Gasiewski, NOAA Environmental Technology Laboratory (ETL), Boulder, CO.
Paul Hwang (EOS PM Project Office) gave the status of the PM project, including the status of the instruments that will fly on the PM-1 spacecraft. The successful completion of the common spacecraft Critical Design Review (CDR), at TRW, in Redondo Beach, CA, allows an on-schedule start of the instruments integration and test, in June 1999. All instruments are on schedule for delivery to TRW. NASDA held a delta CDR for AMSR-E in early July, with the instrument delivery dependent on the outcome out of delta CDR.
Dawn Conway (AMSR-E Software Integrator) showed a revised data products diagram, a Level 2 product software summary, and a listing of all needed ancillary data files. AMSR-E data processing, which will be done at the Marshall Science Investigator-led Processing System (SIPS), will consist of 3 sub-processes. These sub-processes will be running independently of each other, with their eventual merging into one process by launch time.
D. Winebrenner (University of Washington) presented results of his recent research on snow accumulation rates in Antarctica using the 4.5-cm emission: the annual snow deposition in Antarctica is equivalent to 5-7-mm change in the world sea level. He is anxiously awaiting the AMSR-E data, the first 4.5-cm microwave data since SMMR in 1987, to confirm his analysis and to observe any changes that occurred in the region.
The Science Data Validation Plan summary was discussed briefly. The new version of the plan will be online in August.
A. Gasiewski (NOAA ETL, Boulder, CO) presented the status of the PSR, the instrument that he designed while a professor at Georgia Tech. He also showed some preliminary data taken with the instrument over the Labrador Sea in March 1997. Since that flight, Gasiewski has replaced the warm calibration target, and will fly the PSR on the DC-8 during CAMEX-3, out of Patrick AFB.
. A. Chang (Goddard Space Flight Center) discussed a new approach for the snow retrieval algorithm. He plans to use a 'snowpack ensemble' approach where ensemble brightness temperature is generated based on snowpack profiles (depth, density, grain size, and temperature). The processing will be done on a gridded data set (Azimuthal Equal-area with a good aspect ratio: 1:1 at 90° and 2:1 at the equator), mainly for easy matching with the ancillary data that will be used.
J. Comiso (Goddard Space Flight Center) focused on the current status of the analysis of the FIRE-3/ER-2 aircraft validation program over the Arctic. There were eleven ER-2 overflights during the May/June period, nine of which made it all the way to the Surface Heat Budget of the Arctic Ocean (SHEBA) camp (that included a research ship), where scientists have been collecting surface and atmospheric data since October 1997. Preliminary analysis shows good spatial coherence of the MODIS and the Advanced Microwave Precipitation Radiometer (AMPR) images. The time periods of good data are also coincident with rapidly changing ice surface characteristics due to the onset of spring melt as revealed by SSM/I data. The repeat track scheme was shown to have been successfully implemented thereby enabling the collection of coregistered vertically and horizontally polarized data from AMPR. Both data sets (MODIS and AMPR) will be used for testing and validating the sea-ice-retrieval algorithms.
D. Cavalieri (Goddard Space Flight Center) presented a modified version of the NASA team sea-ice algorithm developed in collaboration with Thorsten Markus, an AMSR-E Team research associate at Goddard. The modified NASA Team algorithm makes use of three radiance ratios: the polarization ratio (PR) at 19 GHz and the gradient ratios (GR) between 37 GHz and 19 GHz for both vertical (V) and horizontal (H) polarizations. This modification maintains the temperature independence, the spatial resolution, and the dynamic range of the original NASA Team algorithm. The purpose of the modification was to eliminate the low-ice-concentration bias found in some areas around Antarctica.
T. Wilheit (Texas A&M University) had one of his students, J. Huang, present his research. Huang found that TRMM observations fit the Wilheit et al. (1977) precipitation model after some modifications to the freezing levels and the rain drop distribution are made.
R. Ferraro (NOAA/NESDIS) reported on his activities involving rain-over-land retrieval algorithm. The main issue is the transition of the operational SSM/I land retrieval algorithm into the GPROF framework. (GPROF is the Goddard PROFiling algorithm for instantaneous rain retrieval and has been used for both SSM/I and TRMM.) A new database of hydrometeor profiles and corresponding forward-calculated brightness temperatures which match the empirically derived database of radar and SSM/I measurements can be inserted into GPROF. This data set is based on cloud model simulations and radiative transfer calculations. In this way, both retrieval methods should be virtually identical. However, the advantage of the GPROF scheme is that the hydrometeor profiles can be rederived for a variety of sensor configurations (e.g., frequency, footprint size, etc.). It is hoped that in AMSR-E era, with higher spatial resolution and more channels than SSM/I, hydrometeor profiles can be retrieved over land. Ferraro showed a case study where he compares the new "model" database retrieval output with results from the operational SSM/I retrieval, NEXRAD radar, and current GPROF retrieval. The model and SSM/I retrievals are very close, and show a greater dynamic range than the current GPROF.
E. Njoku (Jet Propulsion Laboratory) presented details on his Level 2 land products retrieval algorithm. Inputs coming from the AMSR-E Main program are, besides the Level 2A AMSR-E science data, sets of static and dynamic ancillary data and a set of program parameters. The Level 2A data are processed first by a gridding module (quality-controlled and registered to the 25-km global cylindrical Equal Area SSM/I Earth (EASE) grid), and then together with the ancillary data, by the Land Surface Parameter Retrieval module. After the iterative retrieval processing and the generation of flags, the output HDF objects are created and the output is sent to the Main program. Njoku also showed some comparisons between Scanning Multichannel Microwave Radiometer (SMMR) and National Centers for Environmental Prediction (NCEP) data.
F. Wentz (Remote Sensing Systems) has been analyzing TMI data to gain insight into its differences with SSM/I. One astonishing finding was the difference in the brightness temperatures measured by the two instruments over the same area: 11.6 K for the V-pol and 8.5 K for the H-pol. TRMM had executed a maneuver where the spacecraft was turned upside down, and the TMI reflector was actually observing cold sky. Wentz found differences of the order of 10 K to 13 K between the observations of the cold sky with the main reflector and with the cold calibration target (cold sky mirror). Two other notes of interest were RFI "hot spots" detected with the 10.7-GHz channel and the diminishing of the El Niño shown by monthly SST maps created from TMI data.
C. Kummerow (Goddard Space Flight Center) tested his TMI rainfall retrieval algorithm and compared it to the output of the TRMM Precipitation Radar and SSM/I. After the additional constraint of "distance to convection" was added to the retrieval algorithm, a significant improvement was achieved in the stratiform rainfall retrievals. Another improvement made to the model in the retrieval algorithm was the decrease of the Marshall-Palmer (M-P) hydrometeor profile intercept (which puts more small droplets into the hydrometeor profile), which brought the retrieval much closer to the observations.
The next meeting will be a Joint AMSR Science team meeting in November, in Tokyo, Japan.