In order to improve the precision of hydrometeorological observation data acquisition, the error calibration was needed for the data parameters obtained from the observation equipment. In this paper, three kinds of error calibration algorithm including piecewise linearization, least squares curve fitting and three polynomial of spline function were compared. Finally, three polynomial fitting of spline function method, which could pass through all the error calibration points and have smooth and continuous curve near the calibration points, was determined to be used for error calibration of the observation data. The experimental comparative analysis shows that three spline function fitting error calibration method can reduce the data measurement error and ensure the observation equipment to obtain a more accurate and reliable data parameter.