Alternative forms of moisture information in 4-D data assimilation

Global data analysis procedures were developed to perform data assimilation for observed geopotential heights, wind components, and relative humidity. This report describes in detail the design of these computer codes and their implementation in conjunction with a 6h global spectral forecast model (GSM) and normal mode initialization. The three procedures together form the AFGL Global Data Assimilation System (AGDAS) which was used in a series of data assimilation experiments to assess the impact of several alternative sources of humidity information used in the relative humidity analysis. Each alternative humidity source was investigated as to reliability of information content in a collocation study against rawinsonde relative humidity. Satellite moisture retrievals, surface weather observations, and AFGWC 3DNEPH cloud amounts were used to deduce upper air relative humidity. The latter two were found in sufficiently good agreement with rawinsonde relative humidity to permit their use in the data assimilation. The 3DNEPH inferred humidities, and to a lesser extent the surface weather observation inferred humidities, resulted in an improved analysis over maintaining the first guess when both were compared with verifying observations. Finally, the sensitivity of the AFGL GSM to initial humidity states was tested by initializing a series of 48h forecasts from several humidity fields.