基于SPOT-6遥感影像的近海水深反演

水深反演对于浅海地形调查、海岸带保护和开发具有重要的意义。本文选取南海东锣岛海域为研究区,采用SPOT-6 6 m高分辨率数据,基于SPOT-6多光谱数据和水深实测值分别建立蓝-绿、蓝-红等6个波段比值模型。结果显示,绿-红波段的比值模型精度最高,R²值达到0.706 4,这对于反演热带海洋地区水深具有一定借鉴意义。基于该模型的反演结果对不同水深范围内的平均相对误差进行比较,结果表明:0~5 m水深范围反演误差较高,主要由于该范围水体情况较复杂,5～10 m 水深范围内平均相对误差最小为13.62%,而随着水深的增加,反演的误差增大,分析误差的主要来源是海水中的悬浮颗粒物、黄色物质、叶绿素质量浓度等方面的影响。

Offshore bathymetry retrieval from SPOT-6 image

Retrieval of offshore bathymetry has important significance in shallow topographic survey, coastal protection and development. In this study, Dongluodao waters in South China Sea was taken as a typical study area, using the SPOT-6 6 m high resolution data, six band ratio model such as blue-green and blue-red were set up respectively based on SPOT-6 multi-spectral data and depth measured values. The results show that green-red band ratio model has the highest accuracy and the R² value reaches 0.706 4. This has certain reference significance for retrieval of offshore bathymetry in tropical oceans. Comparing with the average relative error of the different depths range based on the above model, the conclusion is that the depth range from 0 to 5 m has the higher average relative error, mainly because the conditions are complex. The depth range from 5 to 10 m has a minimum average relative error of 13.62%. Accretion of the average relative error with the increasing of the depth, while SPM, CDOM and chlorophyll concentration are the main causes of this phenomenon.