Application of Nonparametric Methods in Short-Range Precipitation Forecasting

Short-range precipitation forecasting plays a key role in developing public affairs. Seasonal autoregressive integrated moving average (ARIMA), a classic parametric modeling approach to time series, and nonparametric regression models have been proposed as well suited for application to short-range precipitation forecasting. In this paper, the method of the k-nearest neighbor estimation in the nonparametric regression is discussed, and this method is used to establish the day-by-day rainfall forecast of southeastern of Guangxi during the period from May to June. Results show that forecasts from the nonparametric regression scheme are high stability, with good prospects in operational weather forecast.