Intercomparison of the Upper Tropospheric Retrievals from TOVS

The TIROS Operational Vertical Sounder (TOVS) continues to be our most important tool for examining the upper-tropospheric humidity (UTH) and upper-tropospheric water vapor (UTWV). TOVS radiances are used in a variety of retrieval algorithms to produce the global distribution of water vapor over all surface types and in multiple layers. This paper will present the differences in UTWV for many of the most widely used TOVS water vapor datasets. Starting from the same identical TIROS Operational Vertical Sounder (TOVS) radiances, retrievals of atmospheric water vapor are performed by a number of international groups. The scientific uses for these retrievals broadly span the atmospheric sciences at all time and space scales. Some of the most common uses are: improving operational forecasting, addressing basic atmospheric radiative transfer, closing the energy budget, and most recently, efforts to understand the effects of water vapor in the upper troposphere.

The problem under investigation here, is that the retrieval methodology could cause the resulting water vapor amounts to vary greatly, thus influencing the results from scientific investigations such as those listed above. While all layers of water vapor could be examined, this study emphasizes only the upper tropospheric water vapor (UTWV) retrievals from the NOAA operational retrievals (Reale et al., 1989), the TOVS Pathfinder Path A (Susskind et al.,1997), TOVS Pathfinder Path B (Chedin et al.,1994), a new retrieval method from Engelen and Stephens (1999), and also model initialization data from the NCEP/NCAR (Kalnay et al., 1996) and ECMWF Reanalyses.