|Title||Precipitation correction in the ERA-40 reanalysis.|
|Series/Collection||ERA-40 Project Report Series|
|Authors||Troccoli, A, Kållberg, PW|
|Event Series/Collection||ERA Report|
|Place of publication||Shinfield Park, Reading|
The excessive precipitation in the tropical regions, particularly after 1991, is seen as the most serious drawback of the ERA-40 reanalyses. This drawback stems from weaknesses of the humidity scheme utilised in the assimilation system. Motivated by the need to use ERA-40 precipitation fields to force ocean model integrations, a post-processing procedure is applied to the oceanic precipitation fields in order to compensate for this shortcoming. A simple two-step procedure is adopted for this purpose. First, a pseudo-climatology is generated by blending one particular ERA-40 period (i.e., the 1964-1972, which is one of the least affected by the humidity scheme) with precipitation observations. Second, the remaining ERA-40 periods are corrected so as to minimise the distance to the pseudo-climatology, with the constraint that the water budget should be zero. Since the ERA-40 mean evaporation field is fairly constant, the imposed water budget constraint implies that the mean corrected precipitation is very similar for each of the five periods in which ERA-40 was processed. Comparisons with precipitation estimates show a considerable improvement of the corrected precipitation field.