A diagnostic study on atmospheric moisture budget over the continental United States for wet and dry years

The variations in atmospheric moisture budget components over the continental United States are investigated, using the diagnostic method (e.g., Rasmusson 1968, Berbery and Rasmusson, 1999) and the NCEP/NCAR reanalysis (i.e., the windspeeds, specific humidity and pressure) and CMAP (CPC Merged Analysis of Precipitation) data in a wet (1997) and a dry year (2001).

The results show that the area-averaged monthly evaporation estimates range from 0.22 mm/d in February to 2.7 mm/d in August, with monthly precipitations from 1.3 mm/d in February to 2.0 mm/d in July and water vapor flux divergences from -1.4 mm/d in October to 0.7mm/d in August over the continental United States for 1997. As also shown in previous studies, the evaporation estimates are strongly affected by the water vapor flux divergence estimates, and therefore, the variations of evaporation display a very high correlation with those of the water vapor flux divergence, rather than those of precipitation. Besides, the evaporation generally exceed precipitation in the warm season (e.g., June, July, and August), while it is lower than precipitation in the cold season.

In addition, for some particular regions (e.g., the Great Plain, and the core zone of North American monsoon) the range of the 1997 monthly evaporation estimates is larger than that for the entire continental United States. Besides, the values of evaporation estimates might be below zero for some months of the cold season (as also reported by other studies). Generally, large as the resolution of the analyses (i.e., 2.5o lat x 2.5o lon), is, water vapor flux divergence from these data might not be represented as higher than realistic. Therefore, the negative sign of the estimates might imply the underestimates of CMAP in some regions.

Further inspection to the results for the dry year (i.e., 2001) exhibits the similar behavior in variations of the atmospheric moisture budget components to those for the wet year (1997), though the magnitudes in the variations are different. All these demonstrate the consistency for estimating the atmospheric moisture budget components for the two climatologically different years.