海洋底质类型图矢量化智能制作技术研究  

Study of intelligent vector mapping technique of marine sediment types

作　　者：王风帆 孔敏[1] 余佳[1] 韩璐遥 WANG Fengfan;KONG Min;YU Jia;HAN Luyao(National Marine Data and Information Service,Tianjin 300171,China)

机构地区：[1]国家海洋信息中心,天津300171

出　　处：《海洋信息技术与应用》2023年第4期212-217,共6页JOURNAL OF MARINE INFORMATION TECHNOLOGY AND APPLICATION

基　　金：国家重点研发计划(2022YFC2803600,2021YFC3101600)。

摘　　要：海洋底质类型图是海洋学研究重要的基础图件之一。针对现有海洋底质类型图制作方法存在的问题,本文利用随机森林算法,对作图数据网格点处的底质类型进行预测,并基于ArcGIS矢量操作功能改进了底质类型面要素生成与平滑方法,解决了常规成图方法存在的锯齿状边界、后期修改不方便等问题,开发了海洋底质类型图矢量化智能制作技术。利用渤海区域沉积物粒度数据开展应用试验,结果表明,在缺乏沉积物组分信息的情况下,该方法可以获得较好的制图结果,具有一定的实用价值,是对传统底质类型图制作方法的补充与改进。Marine sediment type map is one of the important basic maps for oceanographic research.To solve some problems of existing approaches for sediment types mapping,we applied the random forest algorithm to predict the sediment types of map grid points,and improved the iso-surface generation method based on the vector operation function for ArcGIS,which can solve the problems such as jagged boundaries and inconvenient post-modification in conventional mapping methods and makes it easy to modify later.Finally,we built an intelligent vector mapping technique for marine sediment types,and performed the experiment with sediment grain size data of Bohai Sea.The result shows that this technique could produce a good sediment types map when sediment component information is lacking.It has practical values and is a supplement and improvement for the conventional sediment types mapping methods.

关 键 词：底质类型 类型预测 随机森林 边界平滑 平面分布图 

分 类 号：P717[天文地球—海洋科学] P285.7