Two important implementations, directly related to the prediction of atmospheric humidity fields, have been introduced during in the recent years in the weather forecast model and its data assimilation system at CPTEC/INPE. The first is the inclusion of Integrated Water Vapor (IWV) data in the assimilation scheme. The second is the improvement of the weather forecast model along with an increase in the grid resolution. The goal of this paper is to characterize the impact of these implementations in the prediction of humidity fields. To achieve this goal four experiments were conducted using the T126L28 (without the improvement) and T213L42 (with the improvement) versions of the CPTEC's global model, with and without assimilation of IWV. The results show that the impact of IWV assimilation is larger in the first hours of forecast. On the other hand, the larger impacts produced by the improvements in the weather forecast model occur from the 60th h of forecast onward. Consequently the combined impact of both implementations represent an improvement throughout the integration period or the period at which predictions of IWV present an acceptable skill (above 60%) were extended in up to 25 hours over the global tropical regions and up to 18 hours over South America.
Data assimilation; IWV; PSAS; AIRS-AMSU; NWP
