| Abstract: |
The inter-satellite radiometric calibration of microwave radiometers, known as XCAL, is an important part of NASA’s Global Precipitation Measurements (GPM) program. The constellation of cooperative satellite radiometers is calibrated against a single radiometer operating on a low earth (non-sun synchronous) orbit. Conventional calibration involves XCAL at the cold such as over ocean and, warm, traditionally over tropical rain forest, ends of the scene brightness temperatures. The two-point calibration are connected using a linear regression. Unfortunately, this ignores the possibility of non-linear radiometer transfer function, which is often important.
This paper describes a new method to identify large-scale homogeneous desert scenes as possible third calibration targets with brightness temperatures that fall between the conventional cold and warm earth scenes. The prototype “ShowMask” procedure is developed over continental Australia using 10, 18 and 37 GHz Vertical and Horizontal polarized brightness temperatures from GPM Microwave Imager (GMI) data. The surface classification technique mostly used brightness temperature diurnal variation, second stokes parameter and spatial homogeneity of the scene as key parameters for identifying suitable targets.
This paper will focus on extending the prototype (Australia) tool to global scale and quantitative validation of the mask using surface soil moisture and surface temperature data from Soil Moisture Active Passive (SMAP) observations and also existing latest surface classification map. Further, we will apply the mask to examine intercalibration results over these scenes using GMI-TMI and GMI-AMSR-E matched data pairs. Initial results over Australia indicated a scene temperature dependence of calibration factor, and we will investigate it in more detail in this article.
|
