To reduce the deviation between radar rainfall field and surface rainfall observations, this paper proposes calibrating the radar rainfall field with surface microwave links including the variational calibration method, the Kalman filter calibration method, the mean calibration method and the Kriging calibration method. The rainfall rates retrieved by two microwave links are used to calibrate the Sband radar rainfall field in two precipitation cases of different types. The calibration results are then compared with the measurements of rain gauges. The conclusions are as follows. Firstly, all the four calibration methods are proved effective to reduce the bias between the radarbased rainfall estimates and the gauge measured rainfall. The problems of the underestimation of heavy precipitation in precipitation Case Ⅰ and the overestimation of weak precipitation in Case Ⅱ are both partly solved. The statistical errors including mean absolute error (MAE)， mean error (ME) and root mean square error (RMSE) are all significantly lowered after calibration. The improvement degrees of the statistical errors from high to low are ME, RMSE and MAE. Secondly, the effectiveness of the Kriging calibration method is the best among the four methods. The performances of the variational calibration method and the mean calibration method are better than that of Kalman filter calibration method. The Kriging calibration method is the most effective to reduce ME and RMSE and the variation method is most effective for MAE. Thirdly, the Kriging calibration method and the variational calibration method can derive calibration factor fields which vary with time and spatial position, while the mean calibration method and Kalman filter method can only obtain a mean calibration factor for each time. These results suggest that microwave link can be an effective alternative to calibrate the radar rainfall field.