Robust method for the detection of abnormal data in hydrography

Blunder detection is a topic of great interest to hydrographers because und etected blunders significantly distort the observed parameters, e.g., sound ings. Based on an analysis of the characteristic of marine surveying, a rob ust method for the detection of abnormal data (include blunders) in hydrogr aphy is proposed in this paper, which is called the robust interpolation co mparison test based on robust M-estimation by an iterative calculation proc edure. Some questions involved in the implement of the suggested method are discussed in detail. Compared to the existing methods, the new method has more strong capacity of locating abnormal data. A simulation study and an a ctual numerical example for the process of multibeam soundings are given to test the performance of the proposed method. The results have illustrated the effectiveness of the method in the detection and identification of mult iple blunders. The use of the new method will play an important role in imp roving the quality and reliability of marine measurements in our country.