Automated measurement of sand dune migration using multi-temporal lidar data and GIS

摘要

Remote sensing has been used for coastal and desert sand dune studies over the past four decades, yet few methods have been developed for automated detection and measurement of dune migration directions and migration rates in large dune fields. Using high-resolution, high-accuracy, and multi-temporal light detection and ranging (lidar) data acquired in the White Sands Dune Field in New Mexico (USA) on 24 January 2009 and 6 June 2010, an automated method named Pairs of Source and Target Points (PSTP) was developed in a geographical information system (GIS) environment for automated detection and measurement of dune migration directions and migration rates. As markers for dune movement, dune slip faces were automatically extracted from lidar-derived digital elevation models (DEM), based on the range of the angle of repose, and converted into source lines and target lines through vectorization. Random target points were then generated on target lines and used to search for source points on source lines to form PSTP, thereby obtaining source direction, migration distance, and migration rate for each target point. Continuous raster data sets for dune migration rates were also created through spatial interpolation and point statistics to show dune field scale spatial patterns of dune migration rates. A total of 5936 PSTP were identified, producing dune migration directions and migration rates at 5936 locations in the study area of 9kmx2.4km. Histogram analysis revealed that a majority of the 3025 target points with source direction in the range 225 degrees-285 degrees (direction of prevailing winds) have a migration rate of 4-7myear(-1) and an average migration rate of 5.56myear(-1). The applicability of the PSTP method in special cases, and its potential as a generic method for change detection and measurement, are also discussed. The study obtained important results both in methodology development and in the study area.