Rain is one of the climate variables are particularly concerned by the impact negatively on people's lives. In the season-term forecasts predict good rain will contribute to providing information for many problems in various fields, such as regulating reservoirs, predict droughts, crop restructuring, ... Show This problem-term seasonal rainfall forecast by the numerical model is an inevitable trend. However, due to the complex nature of the phenomenon, the rain forecast model in general is still far to meet actual requirements, including the prediction phase, quantitative. One approach to improve the prediction of rainfall patterns are sought correction of rainfall patterns.
In this study, based on the sequence data of rainfall simulation model was interpolated for you station location and observed rainfall data from 132 meteorological stations Vietnam period 1981-2010, a calibration method rainfall patterns have been built. The method will be implemented in two steps: 1) error correction system model based on the assumption of linear relations; and 2) Edit distributed exponential formula. The results obtained from the data dependency chain saw after adjusting for the quality of the model precipitation has improved significantly with the mean error (ME), absolute error (MAE) as per the wrong number (RMSE) are reduced to a correlation between rainfall and rainfall patterns observed a clear increase.
đang được dịch, vui lòng đợi..