TES atmospheric profile retrieval characterization: an orbit of simulated observations

Abstract

We test the accuracy of our error analysis and retrieval performance by examining retrievals over an orbits' worth of simulated data covering a variety of atmospheric conditions. The use of simulated data allows validation of the error analysis and retrieval algorithm by comparisons to the true values. To demonstrate typical results, two example retrievals are shown, along with associated diagnostic information. Curtain plots display comparisons between the retrieved results, the true values, and the initial guesses. The results show that the Tropospheric Emission Spectrometer (TES) retrieval algorithm is robust under a variety of atmospheric conditions, that TES can improve on the a priori for nadir species Tatim H/sub 2/O, O/sub 3/, and CO, and that the predicted errors match well with the actual retrieved errors. The target scenes (nadir, ocean, cloud-free) simulate conditions that are most easily validated with real data, and comparisons of on orbit results can be made with this baseline.