Study On Data Processing And Quality Control In Geologic Hazard

Geological hazard appears complicated as a result of multiple aspects in social and economic condition. A great quantity of cases and impact factors should be considered and analyzed. In GIS(Geo-information System), data plays the role of the kernel as well as the foundation of related work. Providing the basis of secondary development and judgment criteria, data quality is closely linked with the reliability of GIS work. In order to promote long-term development and related researches, GIS productions should be of high quality and precision. Data quality includes data error and data accuracy. The source of error is data itself as well as the processing of data, which can be divided into systematic error and non-systematic error. Attribute data and spatial data possess different characters. With different error occurs in different characters, methods of correction are diverse.Research subject about data quality and control has been put forward. Based on data of Pulan county geological hazard, Tibet, methods about that processing geological hazard data quality inspection and optimizing is quested in order to improve the quality.According to DZ, data quality inspection includes integrity, veracity, logic consistency and so on. Geological hazard data is divided into two piles, attribute data and spatial data. Attribute data reveals the status of collections and survey, while spatial data reveals geographical position, geographic and geomorphic. The main study contents include 3 aspects.(1)According to the basic theory of geological hazard survey and DZ, attribute data accuracy and integrity are inspected. Because of different types, different quality detection standards are selected to processing inspection. Supplementing missing data, deleting redundant data, and selecting the data by AHP method;(2)To the spatial data, there has many quality problems, for example the missing contour lines, geography registration errors. On the basis of TIN and IDW methods, contour lines are supplemented reasonably. Remote sensing images are classified by ISODATA unsupervised classification method. Control points are selected in the course of shape feature theory, to georeference images;(3)Banding attribute data and spatial data together to processing associated examine, handling with the results, and scheduling the next steps of work.