The Application of Parallel Algorithm Technology for Watershed Terrain Analysis Based on DEM

In this paper, Arcgis software is used to calculate the three-dimensional vector of water environment at multiple stations, and the original data space is divided into blocks to achieve the comprehensive evaluation of watershed terrain based on DEM (Dynamic Effect Model). In this method, an artificial neural network is first used to classify rivers at each cross-section. Next, by comparing the weights corresponding to different algorithms, the optimal thresholds for water depth, depth, and slope were determined for each time period. Finally, combining the least squares support vector machine Law of Return method and interpolation analysis technology, the comprehensive evaluation result of watershed terrain based on DEM is realized, which is more accurate than the traditional single factor quantitative calculation. The results are as follows: before parallel algorithm optimization, the accuracy of soil erosion risk assessment is 0.3, and after optimization, its accuracy is improved to 0.6. Similarly, the accuracy of sediment yield assessment was 0.4 before parallel algorithm optimization, but after optimization, it reached 0.7. In addition, this technology can significantly improve the stability of computational data.