Based on the three different types of precipitation data of Tianjin and Doppler radar observational data, using the radar rainfall estimation grid point field derived from the all valid raingauge calibration as a “true” value, the accuracy of radar rainfall estimates by 14 different raingauge density calibrations is studied, and the contribution of one raingauge to radar rainfall estimates is also analyzed. The conclusion can be drawn as follows: (1) Radar rainfall estimation error is bigger when the raingauge calibration density is lower, then estimation results are not satisfactory. However the estimation accuracy of radar is continuously improved and estimation error is significantly reduced, with the raingauge calibration density increased. (2) Calibration on different types of precipitation needs different raingauge densities, which is related to the nature of precipitation. When the precision of radar estimation is equal to the precision of the “true” value, the raingauge maximum density on calibration radar is required to be one raingauge per 121 km2. (3) Any increased raingauge site, whose precipitation is zero, has no impacts on radar rainfall estimation. When it is not zero, the estimation error would appear around the site (over or under estimate), and the deviation of estimation is closely related to the average precipitation and the distribution of raingauge density around the raingauge site. (4) When randomly increased (decreased) precipitation of single raingauge site, the radar would be over or under estimated the precipitation around the raingauge site, and the degree of over or under estimates and the area of bias are related to the density of the surrounding raingauges and the changes in the size of precipitation, which was unrelated to the type of precipitation and either unrelated to the variation of precipitation over time.