地表液态水调查数据质量控制关键技术研究

Research on Key Technologies for Data Quality Control in Surface Liquid Water Surveys

摘要: 地表液态水储存量调查对基础空间数据质量要求较高，其中水陆一体数字高程模型(DEM)是水体储量计算的重要基础数据。针对水陆一体DEM在生产和融合过程中易出现噪声与异常高程值的问题，本文研究了一种基于多尺度滤波的DEM粗差检测方法。该方法综合采用均值滤波、中值滤波、高斯滤波、双边滤波和形态学滤波，对DEM数据进行平滑处理，并通过残差分析与阈值判别实现高程异常的自动识别。在此基础上，基于Python语言开发了DEM粗差检测自动化工具，实现了DEM数据的批量处理与结果可视化。以辽宁省地表液态水储存量调查中的水陆一体DEM数据为例开展实验验证，结果表明，该方法能够有效识别DEM中的高程异常，整体检测准确率稳定在95%以上，处理效率较传统方法明显提高。

Abstract: The investigation of surface liquid water storage has high requirements for the quality of basic spatial data, among which the water-land integrated Digital Elevation Model (DEM) is important fundamental data for water volume calculation. Aiming at the problem that noise and abnormal elevation values are prone to occur in the production and fusion process of water-land integrated DEM, this paper proposes a DEM gross error detection method based on multi-scale filtering. The method comprehensively adopts mean filtering, median filtering, Gaussian filtering, bilateral filtering and morphological filtering to smooth DEM data, and realizes automatic identification of elevation anomalies through residual analysis and threshold discrimination. On this basis, an automatic tool for DEM gross error detection is developed based on Python, realizing batch processing of DEM data and visualization of results. Experiments are carried out using water-land integrated DEM data from the surface liquid water storage survey in Liaoning Province for verification. The results show that the proposed method can effectively identify elevation anomalies in DEM, with the overall detection accuracy stably above 95%, and the processing efficiency is significantly improved compared with traditional methods.