In this paper, green-channel waveforms obtained with Mapper5000 which is a new airborne LiDAR bathymetry (ALB) system are analyzed. A waveform comparison method is proposed to extract the characteristics of the waveforms. By deducing the theoretical distribution of the surface return intensity, the causes of the errors of the surface signal detection are determined and classified into two types. Then the main influence factors of the errors are statistically discussed. The results show that the detection errors from the extremely shallow water depend on the depth, while the errors caused by the weak surface return are mainly determined by scan angle, sensor altitude and clarity of the water.