Algorithm Theoretical Basis Document (ATBD) Prepared by Remote Sensing Systems
Due to diffraction, radiometers of differing frequencies using a common antenna do not generally produce equivalent gain patterns on the Earth surface. Consequently, direct comparison of such observations is complicated by the fact that the measurements do not describe identical locations.
The Level 2A algorithm will alleviate this problem by producing several spatially consistent data sets, corresponding to the footprint sizes of the 6.9, 10.7, 18.7, 36.5, and 89 GHz observations.
These five sets of antenna patterns are subsequently referred to as resolutions 1 though 5 of the Level 2A data set, corresponding to footprint sizes of approximately 58, 37, 21, 11, and 5 km respectively. Observations will be produced at spatial intervals of approximately 10 km for the next four lowest resolution data sets, and 5 km for the highest resolution set.
Throughout the remainder of this document, brightness temperature measurements derived directly from the Level 1A stage of processing will be referred to as “actual observations.” Those brightness temperatures that are the result of a linear combination of actual observations will be identified as “constructed observations,” or “effective observations.” The gain patterns of the effective observations closely match actual antenna gain patterns while reducing the effect of measurement noise through averaging.
For each Level 2A observation within a single scan of the instrument, a set of coefficients describes the relative weights of the neighboring actual observations that are combined to produce the effective observation. These sets of coefficients are unique for every
position along the instrument scan, but do not vary from scan to scan.
For More: Remote Sensing Systems