A parametric optimal estimation retrieval of the non-precipitating parameters over the global oceans
Date
2006
Authors
Elsaesser, Gregory S., author
Kummerow, Chris, advisor
Reising, Steven C., committee member
Randall, David, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
There are a multitude of spacebome microwave sensors in orbit, including the TRMM Microwave Imager (TMI), the Special Sensor Microwave/lmager (SSM/I) onboard the DMSP satellites, the Advanced Microwave Scanning Radiometer - Earth Observing System (AMSR-E), SSMIS, WINDSAT, and others. Future missions, such as the planned Global Precipitation Measurement (GPM) Mission, will incorporate additional spacebome microwave sensors. The need for consistent geophysical parameter retrievals among an ever-increasing number of microwave sensors requires the development of a physical retrieval scheme independent of any particular sensor and flexible enough so that future microwave sensors can be added with relative ease. To this end, we attempt to develop a parametric retrieval algorithm currently applicable to the non-precipitating atmosphere with the goal of having consistent non-precipitating geophysical parameter products. An algorithm of this nature makes is easier to merge separate products, which, when combined, would allow for additional global sampling or longer time series of the retrieved global geophysical parameters for climate purposes. This algorithm is currently applied to TMI, SSM/I and AMSR-E with results that are comparable to other independent microwave retrievals of the non-precipitating parameters designed for specific sensors. The physical retrieval is developed within the optimal estimation framework. The development of the retrieval within this framework ensures that the simulated radiances corresponding to the retrieved geophysical parameters will always agree with observed radiances regardless of the sensor being used. Furthermore, a framework of this nature allows one to easily add additional physics to describe radiation propagation through raining scenes, thus allowing for the merger of cloud and precipitation retrievals, if so desired. Additionally, optimal estimation provides error estimates on the retrieval, a product often not available in other algorithms, information on potential forward model/sensor biases, and a number of useful diagnostics providing information on the validity and significance of the retrieval (such as Chi-Square, indicative of the general "fit" between the model and observations and the A-Matrix, indicating the sensitivity of the model to a change in the geophysical parameters). There is an expected global response of these diagnostics based on the scene being observed, such as in the case of a raining scene. Fortunately, since TRMM has a precipitation radar (TRMM PR) in addition to a radiometer (TMI) flying on-board, the expected response of the retrieval diagnostics to rainfall can be evaluated. It is shown that a potentially powerful rainfall screen can then be developed for use in passive microwave rainfall and cloud property retrieval algorithms with the possibility of discriminating between precipitating and nonprecipitating scenes, and further indicating the possible contamination of rainfall in cloud liquid water path microwave retrievals.
Description
Covers not scanned.
Print version deaccessioned 2021.
Print version deaccessioned 2021.