A Novel Method for Automatic Extracting Loess Positive and Negative Terrains Based on Grid DEMs

The Loess Positive and Negative Terrains (P-N terrains), which are separated by loess shoulder-lines obviously, are crucial parameters in constructing the mechanism-processing models of the loess landform drainage systems. This paper proposed a novel algorithm for automatic extracting loess P-N terrains by using of 5m resolution Digital Elevation Models. Three main procedures are designed in the algorithm. The first stage is to detecting out grid cells which at the location of shoulder-lines by comparing the slope gradient up and down the hill slope. Slope gradient variance 20°, 25°and 30°are used as the thresholds in Loess Platform area, Loess Ridge-hill area and Loess Hill area respectively in the extraction model. The second stage is to expanding shoulder-lines´ cells directionally by considering the spatial direction of the local hill slope aspect and the trend of the shoulder-lines. This step is helpful for deriving more consecutive and detailed shoulder-lines on DEMs. The third stage is to generating Positive terrain areas by combination of the upstream contribution areas which controlled by the shoulder-lines grid cell points derived in the step above. The Negative terrain areas are remnant regions deducting from Positive terrain areas. Validation tests are took out by contrasting the auto-extraction results from DEMs and the delineating results from 1m resolution DOMs in six drainages which belonging to typical loess landforms. Results show that the main advantages of this approach are high accuracy, lower demands on manual intervention and ready availability of required data for many regions on the Loess Plateau.