Instruments: CoSSIR/CoSMIR
 |
Figure 1: Typical multi-channel brightness temperature images of ice clouds acquired by CoSSIR near Florida on July 1, 2002.
+ Larger image
(opens in a new window)
|
Currently, there are two major areas of endeavors. The first one is the development and improvements of the airborne millimeter-wave/submillimeter-wave radiometers, the Conical Scanning Millimeter-wave Imaging Radiometer (CoSMIR, 50-183 GHz) and Compact Scanning Submillimeter-wave Imaging Radiometer (CoSSIR, 183-874 GHz). The scope of the work is not limited to the hardware development and improvements, but also includes data acquisition in various field campaigns and ensuing data analysis. Both systems can acquire conical scanning as well as across-track scanning data sets simultaneously. CoSMIR is mainly used for the calibration/validation of the SSMIS (Special Sensor Microwave Imager/Sounder), a new-generation orbiting microwave radiometer that serve as the precursor to the CMIS (Conical scanning Microwave Imager/ Sounder) in the future NPOESS satellites. CoSSIR can be used to measure water vapor, snowfall, and the snow cover on the ground. CoSSIR is mainly used for the measurements of ice clouds, a key atmospheric parameter that is solely needed for input to the modeling of the earth’s water and energy cycles. Besides ice clouds, the CoSSIR can be used to measure water vapor as well as snowfalls. The measurements from the channels around the two water vapor lines of 183.3 and 380 GHz enhance the water vapor profiling capability that was not available from the previous microwave sensors such as airborne MIR (Millimeter-wave Imaging Radiometer) or AMSU-B (Advanced Microwave Sounding Unit B). The following three figures demonstrate the CoSSIR capability of measuring atmospheric parameters.
 |
Figure 2: Retrievals of particle size and total column of ice clouds.
+ Larger image
|
Figure 1 (above) shows the typical multi-channel brightness temperature images of ice clouds acquired by CoSSIR near Florida on July 1, 2002. Several cells of ice clouds with low brightness temperatures are clearly displayed. This demonstrates that ice clouds strongly scatter submillimeter-wave radiation and that the brightness depressions are generally larger at higher frequencies. CoSSIR measurements like this can be readily used for retrievals of particle size and total column of ice clouds, as demonstrated by the results in
Figure 2 (left). The top two panels of this figure give the ice water path and the median volume equivalent sphere diameter, respectively. These two parameters together are used to calculate the integrated backscattering, which is compared with that measured by another instrument (Cloud Radar System, or CRS, on board the same aircraft) in the bottom panel of the figure. The excellent agreement between the CoSSIR retrieved and CRS measured integrated backscattering implies the soundness of the CoSSIR measurements and subsequent retrieval approach.
 |
Figure 3: An example of water vapor profiling with CoSSIR measurements.
+ Larger image
(opens in a new window)
|
Figure 3 (right) provides an example of water vapor profiling with CoSSIR measurements. The data set was obtained during the transit flight from Houston, Texas to Costa Rica on January 14, 2006. Eight channels of measurements at 183.3±1, 183.3±3, 183.3±6.6, 220, 380±0.8, 380±1.8, 380±3.3, and 380±6.2 GHz are used in the retrieval. The past efforts of water vapor profiling were made at frequencies £ 183 GHz, and retrievals were limited to altitudes £ 10 km because of lack of measurement sensitivity. The addition of the measurements near the strong water line at 380 GHz extends the profiling capability to about 13 km. The figure displays the retrieved water vapor mixing ratio at seven different altitudes, together with that measured by the Meteorological Measurement System (MMS) in proximity to the aircraft. Some correlation between the MMS measured and CoSSIR retrieved mixing ratio is noticeable, although the altitudes of measurements are many km apart.
Another area of work is in the application of atmospheric corrections to the retrievals of snow water equivalent (SWE) using AMSR-E measurements at 18.7 and 36.5 GHz. The effects atmospheric absorption by oxygen and liquid clouds at these frequencies are not small, even though most widely used algorithms to retrieve SWE depend on the brightness difference between these two channels.
Figure 4 (left) demonstrates such effects for two selected regions with AMSR-E passes, one on the left in Sierra-Nevada, and another on the right centered at Fargo, North Dakota. The top two plots give the SWE directly estimated from the AMSR-E measurements. The middle two plots give the SWE distribution if measured at the ground level. The bottom two plots are the SWE derived from the surface emission alone (i.e., no atmosphere). Clearly, there are substantial differences in SWE values depending on how the measurements are made. Developing a reasonable approach to account for these atmospheric effects will help quantify other noise factors like vegetation cover, and will ultimately lead to more reliable retrievals of SWE. The work can be expanded to coordinate with snowfall estimation from the AMSU-B measurements to bring about a better understanding of the connection between snowfall and snow on the ground.
