Potential of temporal information of high density airborne LiDAR data forinstream flow type classification

Instream habitat mapping is an important task for river management. Remote sensing techniques have been successfully applied to replace the conventional method of habitat mapping due to the disadvantages of conventional method like expansive, labor intensive and time-consuming. Airborne LiDAR (Light Detection And Ranging) combining standard deviation analysis has been proved to be effective for instream habitat mapping in the last research. In this research, we improved the point density of LiDAR data and collected the ground truth data with the aim of GPS for accurate positioning. New methods using temporal information of water surface to discriminate instream flow types are tested and compare to the original method. The result shows that standard deviation of average surface elevation provides the best classification accuracy, and as the LiDAR technique improved, it has a great potential to be a useful tool for instream flow type classification in the future.