Comments, questions on this web page? Send e-mail to: |
Accomplishments in FY041. Proposed activities in FY04
2. Accomplishments and findings in FY042.1 ABI band simulation from current satellite dataABI band studies include the band selection, the spectral coverage and the merit of each spectral band in the weather and environmental applications. 16 bands covering visible (0.47 and 0.64 mm), near IR (0.86, 1.38, 1.61, 2.26 mm) and IR (3.9, 6.19, 6.95, 7.34, 8.5, 9.61, 10.35, 11.2, 12.3 and 13.3 mm) spectral regions have been selected for ABI (http://cimss.ssec.wisc.edu/goes/abi). ABI bands have been simulated using existing high spectral resolution AVIRIS, AIRS, NAST-I, high spatial resolution MODIS, and high temporal resolution MSG data to demonstrate the spectral and spatial characteristics and also provide simultaneous data set of all ABI spectral bands for algorithm testing and development. Figure 1 shows a simulated "16 band" ABI multiple panel image from April 11, 2004 at approximately 13 UTC. This image over France is built from measurements from three separate satellite instruments (MODIS, MET-8 and AIRS). 
Figure 1. A simulated "16 band" ABI multiple panel image from April 11, 2004 at approximately 13 UTC. 2.2 ABI band study for SO2 detectionAn upper level Sulfur Dioxide (SO2) plume from a volcanic eruption has been investigated. The ABI volcanic ash detection will depend on 12 mm data (Schmit et al. 2001), plus 7.34 and 8.5 mm data. The ABI water vapor bands have been shifted to detect upper-level SO2, as well as the mid-level water vapor. Forward calculations with the LBLRTM (left panel, Figure 2) show the sensitivity of the 7.34 and 8.5 mm bands to SO2 for the U.S. standard atmosphere. If the vertical and spatial extents of the plume are sufficient, the ABI should be able to detect medium to large SO2 amounts (greater than 100 Dobson Units (DU)), but not small amounts (less than 50 DU). HES, if they include the 7 mm region, should have more sensitivity than broadband radiometers for detecting SO2 because individual absorption features can be used. ABI bands, simulated from AIRS data, show the potential for depicting upper-level SO2 (right panel, Figure 2). This case is from July 13, 2003 for the eruption of the Soufriere Hills volcano on the island of Montserrat in the eastern Caribbean, images from AIRS data were created to simulate the two ABI bands (13.3 mm and 7.34 mm bands). The resulting BT difference (between 13.3 mm and 7.34 mm) image shows the SO2 plume. This plume agrees well with the Total Ozone Mapping System (TOMS) SO2 observation. Considering the high spatial resolution of ABI, delineation of upper level SO2 is promising. 
Figure 2. The SO2 detection from the BT difference between AIRS 13.3 mm and 7.34 mm channels, as well as the Total Ozone Mapping System (TOMS) SO2 observation. 2.3 ABI band elimination study for cloud maskingThe objective is to study the impact of without some ABI spectral bands
The impacts on various products are studied. Especially, the impact of the ABI band elimination on the cloud detection and masking is studied. Due to the unique information provided by each of the four ABI spectral bands that are being studied for elimination, along with the effects on a host of products, none of these bands should be removed from the ABI. Figure 3 shows the classification masks with all ABI spectral bands (left panel) and without 2.26, 7.4 and 0.47 µm spectral bands (with only 13 ABI bands) (right panel). The circle area has low clouds with all ABI bands. However it is classified as clear with scattered low clouds when the 2.26, 7.4 and 0.47 µm are excluded. MODIS data (MOD021KM.A2002179.1640.004.2003208221115.hdf ) are used in the study. 
Figure 3. The classification masks with all ABI spectral bands (left panel) and without 2.26, 7.4 and 0.47 µm spectral bands (with only 13 ABI bands) (right panel). 2.4 ABI 3.7 µm versus 3.9 µm studies
Figure 4. Classification difference matrix from using 3.9 µm + other ABI bands to using 3.7 µm + other bands. A trade-off study has been done to compare the 3.7 mm and 3.9 mm for ABI shortwave IR band that is very important for the cloud and fire detection. Both temperature and moisture sensitivities of 3.7 mm and 3.9 mm are analyzed. MODIS multi-spectral band data have been used to investigate the surface and cloud type classification difference between using 3.7 mm and 3.9 mm, results show that (1) 3.7 mm and 3.9 mm have the similar temperature and moisture information; (2) 3.7 mm and 3.9 mm have slight differences in day time cloud detection/classification due to the different solar reflectance, while they are almost identical in night time cloud detection/classification; (3) 3.7 µm should be better in cloud detection than the 3.9 µm during daytime, while 3.9 µm should be better than the 3.7 µm on fire detection. Figure 4 shows the classification difference matrix from using 3.9 µm + other ABI bands to using 3.7 µm + other bands. Of-diagonal means the percentage of pixels assigned to a class with 3.9 µm + other ABI bands, now assigned differently to the other class with 3.7 µm + other ABI bands. 3.7 mm detects more low clouds than 3.9 mm in this case. Daytime MODIS data are used. 2.5 ABI temporal resolution studyThe objective is to continue ABI temporal resolution study, and to answer the question on what the ABI temporal resolution we need. Studies include:
Studies show that 5-minute ABI observations are very important to delineate the detailed features and structures of significant weather systems such as storm and convective systems. Movies on ABI temporal resolution study have been included separately in the attached CD. 2.6 Continue trade-off study for HES (spectral coverage, spectral resolution, spatial resolution, signal-to-noise etc.)The objective is to find optimal balance among spectral coverage, spectral resolution, spatial resolution, signal-to-noise, etc. based on the users' requirement and technical requirement, and to establish a linkage between the science requirement and instrument requirement. Accomplishments and findings include
Figure 5. NAST-I Relative Humidity (RH) (%) crosses section at 2 km spatial resolution (upper left panel), 4km resolution (lower left panel), 8km resolution (upper right panel), and 10 km resolution (lower right panel). The date is 26 July 2002 over the ocean near Florida. Figure 5 show the NASTI water vapor retrieval cross-section with various spatial resolutions; higher spatial resolution is needed for depicting the moisture spatial gradients. 
Figure 6. The temperature (left panel) and water vapor relative humidity (RH) (right panel) retrieval rmse from HES LW + LMW, LW + SMW, and LW +LMW+ SMW. Figure 6 shows the temperature (left panel) and water vapor relative humidity (RH) retrieval root mean square error (rmse) from simulated HES radiances for the following configurations: LW + LMW, LW + SMW, and LW +LMW+ SMW. (LW: 650 - 1200 cm-1; LMW: 1200 - 1650 cm-1; SMW: 1650 - 2250 cm-1). 650 global independent retrievals are included in the statistics. Both LMW and SMW are recommended for HES spectral coverage. Spectral resolution and signal-to-noise ratio also determines the vertical resolution of temperature and moisture soundings. Higher spectral resolution and better signal-to-noise ratio correspond to higher vertical resolution. Figure 7 shows the temperature vertical resolutions from HES two-band option (LW+SMW) with reduced noise, nominal noise and double noise, respectively. A reduced noise significantly improves the vertical resolution of temperature retrieval. The vertical resolution study will continue on spectral resolution, spectral coverage, etc. 
Figure 7. The temperature vertical resolutions from HES two-band option (LW+SMW) with reduced noise, nominal noise and double noise, respectively. 2.7 HES/current GOES Sounder simulation study using MM5The objective is to create continuous cube data of atmospheric field for ABI/HES simulation study, especially to demonstrate the capability of high spectral and temporal resolution ABI/HES system over the current GOES imager/sounder system on monitoring the weather evolution. Accomplishments and findings include: Fine spatial resolution (2 - 4 km spatial resolution) cube data from regional model MM5 has been created for HES and current GOES sounder radiance simulation for both clear and cloudy skies during IHOP. The time step for the successive cube data is 30 minutes. Forward and inverse models are created to simulate the radiances and retrievals for HES and current GOES sounder.
Figure 8. 850 hPa true water vapor (upper left), retrievals from HES (upper right), retrievals from the current GOES sounder (lower right), and the differences between HES retrievals and the current GOES sounder retrievals (lower left). Figure 8 shows the 850 hPa true water vapor mixing ratio from MM5 used to simulate HES and current GOES sounder radiances (upper left), water vapor retrievals from simulated HES radiances (upper right), water vapor retrievals from simulated current GOES sounder radiances (lower right), along with the differences between HES retrievals and the current GOES sounder retrievals (lower left panel). This is single time step retrieval from HES and GOES simulated radiances, HES is closer to the truth than current GOES Sounder in term of accuracy as well as spatial features. Movies on HES and current GOES Sounder simulations from have been included separately in the attached CD. 2.8 ABI/HES synergism studiesThe objective is to investigate the better use of data from ABI/HES system, algorithms for products by synergistically using the ABI and HES are developed. MODIS/AIRS on EOS Aqua are used in the study. Accomplishments and findings include:
The advantage of ABI/HES synergism is also studied using MODIS and AIRS data, for example, ABI can help HES in (1) cloud detection, (2) sub-pixel cloud characterization by identifying the cloud phase mask and cloud classification mask, (3) cloud-clearing for partly HES cloudy footprints, (4) improving the retrieval by providing the background and first guess information in the variational approach. Results show that cloud property retrievals from the combination of imager/sounder (ABI/HES) are better than those from either imager (ABI) alone or sounder (HES) alone (Li et al. 2004b). Figure 9 shows an AIRS window image of September 06, 2002 (upper left panel), the 1km MODIS classification mask superimposed to a study area (indicated in upper left panel) for the AIRS sub-pixel cloud characterization (upper right panel), and spectra of AIRS Brightness Temperature (BT) observation (black line), BT calculations with cloud-top pressure (CTP) and effective cloud amount (ECA) from MODIS (green line), CTP and ECA from MODIS+AIRS (blue line), CTP, cloud particle size (CPS) in diameter cloud optical thickness (COT) at 0.55 mm from MODIS+AIRS (red line), respectively in the lower panel, for the indicated AIRS cloudy footprint. Single-layer and mid-level clouds are found by MODIS 1km classification mask over Lake Michigan. They are also ice clouds identified by the 1km MODIS cloud phase mask. Calculation with CTP, CPS and COT from combined MODIS and AIRS fits better the observation for the cloud footprint. 
Figure 9. AIRS sub-pixel cloud characterization using the MODIS 1km classification mask superimposed to the AIRS footprints. The advantage of synergistic use of MODIS and AIRS in cloud property retrieval is also indicated in the lower panel. 3. Peer-reviewed Journal Papers on ABI/HES studies in FY04
4. Selected Proceeding Papers on ABI/HES studies in FY04
5. Select presentations on ABI/HES studies in FY04
6. Acronym list
|