There are numerous methods for data filling in hydrology. However, nearly all of them are based on the correlations with nearby stations, in a general scheme of regionalization. These methods, though robust, fail to function where and when all the near stations are missed-data too. Mashhad synoptic station has annual rainfall data over a 50 years period for 1951 to 2000 and a historic rainfall data before Second World War II from 1893 to 1940. This time series has around 15 years of missed data which can not be filled by the common methods. So, we adopted the techniques of geostatistics and kriging to this long-term time series as an alternative. The data showed a poor correlation at every time lag, so all the semi-variogram models performed nearly equal, yet a high correlation among each others. The results included with polynomial regression fits to different moving average orders, nailed at some reasonable estimates for missed rainfall values.

Keywords