A Coupled-Path Retrieval Algorithm for High Resolution Imagers

Retrieval algorithms for typical space borne cross-track or conically scanning imagers assume that the same atmospheric column is observed along the upwelling and downwelling propagation paths. This assumption is generally valid because of the relative observing geometry of the instrument's horizontal footprint and the freezing level height of the atmosphere. Since the horizontal footprint of typical space borne imagers is much larger than the freezing level, upwelling and downwelling atmospheric columns can be modeled with identical precipitation contributions. This greatly simplifies retrieval algorithms by decoupling the retrieval in each viewing direction from that in every other direction.

There are situations in which the identical atmosphere assumption is not valid. These situations motivate the development of a new coupled-path precipitation retrieval algorithm. If the horizontal footprint is comparable to the freezing level height of the atmosphere, there are potential situations in which the precipitation seen along the upwelling propagation path is dissimilar from that seen along the downwelling propagation path. In this case, the forward radiative transfer model (FRTM) needs to take into account two separate precipitation inputs: one for the upwelling TB and one for the downwelling TB. This added complexity allows for the rain absorption coefficient to be computed separately for the upwelling and downwelling paths. More importantly, it couples the retrieval between viewing directions because, for example, the downwelling atmosphere in one viewing direction can be the upwelling atmosphere in another. Using an iterative least squares estimator, surface wind speeds and rain rates are both retrieved simultaneously at all viewing directions with the coupling between viewing directions fully accounted for by the FRTM. The jacobian used in this coupled-path process represents a weighting function that defines which precipitation observations along the full swath of observations contribute to a single TB.

The coupled-path algorithm is demonstrated using observations made by the Hurricane Imaging Radiometer (HIRad) – an improved version of the Stepped Frequency Microwave Radiometer that is currently under development by NASA, NOAA, the University of Central Florida and the University of Michigan. HIRad's horizontal footprint is similar to that of the freezing level height. Using the coupled-path retrieval algorithm, typical hurricane rain distributions can be accounted for. Flights over Hurricanes Earl and Karl during GRIP (Genesis and Rapid Intensification Processes) provided observations for developing a coupled-path precipitation retrieval algorithm for HIRad. However, other high-resolution imagers could potentially employ this coupled-path algorithm method.