Performance evaluation of different estimation methods for missing rainfall data

There are numerous methods to estimate missing values of which some are used depending on the data type and regional climatic characteristics. In this research, part of the monthly precipitation data in Sarab synoptic station, east Azerbaijan province, Iran was randomly considered missing values. In order to study the effectiveness of various methods to estimate missing data, by seven classic statistical methods and M5 model tree as one of efficient data mining methods, hypothetical missing values were estimated using precipitation data from neighbor station. The results showed that multiple imputation, Delta Learning Rule, and Multivariable Linear Regression (MLR) yield relatively more accurate results with fewer errors. The results also indicate the fact that, developing if-then rules, M5 tree model, as one of modern data mining methods, has been able to give the most accurate results among the mentioned methods with four simple linear relationship and statistical values including correlation coefficient (0.974), Nash-Sutcliffe model efficiency coefficient (0.948), RMSE (5.11), and MAE (4.189). Therefore, taking simple modeling process, functionality, comprehensibility, and high accuracy of this method into account, this method is proposed to estimate monthly precipitation missing values.