The analogy and predictability of the forecasting model error for the precipitation over the mid-lower reaches of the Yangtze River in summer

Abstract

This paper reports an effective method to improve the forecasting level of the numerical model through analogue prediction of errors and correction of the results. The analogy of the precipitation model errors and its predictability are studied for the mid-lower reaches of the Yangtze River in summer time in the perspective of analogy, which exists in the error field in the forecasting numerical model. The content of the analogy is also investigated according to the historical data. It is found that the forecasting errors could be improved remarkably by analogue error prediction over the regions researched in summer time. The forecasting error field is decomposed by EOF, and then the geographic distribution and time coefficient evolution of the first three principal components are analyzed. The prediction of the precipitation could be simplified by analogue forecasting of the principal components separately, and it is more targeted to improve the potential forecasting level. On the basis of the analogy of the forecasting error field, its analogue predictability is defined to measure the predictability of the errors. The analogue predictability of the first three principal components is significantly higher than that of the original field. It has potential applications to precipitation predication by forecasting the error field principal components.