Abstract Eutrophication of inland waters is considered a serious global environmental problem. Satellite remote sensing (RS) has been established as an important source of information to determine the trophic state of inland waters through the retrieval of optically active water quality parameters such as chlorophyll-a (Chl-a). However, the use of RS techniques for assessment of the trophic state of inland waters on a global scale is hindered by the performance of retrieval algorithms over highly dynamic and complex optical properties that characterize many of these systems. In this study, we developed a new RS approach to assess the trophic state of global inland water bodies based on Moderate Resolution Imaging Spectroradiometer (MODIS) imagery and the Forel-Ule index (FUI). First, the FUI was calculated from MODIS data by dividing natural water colour into 21 indices from dark blue to yellowish-brown. Then the relationship between FUI and the trophic state index (TSI) was established based on in-situ measurements and MODIS products. The water-leaving reflectance at 645 nm band was employed to distinguish coloured dissolved organic matter (CDOM)-dominated systems in the FUI-based trophic state assessment. Based on the analysis, the FUI-based trophic state assessment method was developed and applied to assess the trophic states of 2058 large inland water bodies (surface area >25 km2) distributed around the world using MODIS data from the austral and boreal summers of 2012. Our results showed that FUI can be retrieved from MODIS with a considerable accuracy (92.5%, R2 = 0.92) by comparing with concurrent in situ measurements over a wide range of lakes, and the overall accuracy of the FUI-based trophic state assessment method is 80.0% (R2 = 0.75) validated by an independent dataset. Of the global large water bodies considered, oligotrophic large lakes were found to be concentrated in plateau regions in central Asia and southern South America, while eutrophic large lakes were concentrated in central Africa, eastern Asia, and mid-northern and southeast North America.

中文翻译：

---

使用 MODIS 衍生的 Forel-Ule 指数对全球内陆水域的营养状态进行评估

摘要 内陆水域富营养化被认为是一个严重的全球环境问题。卫星遥感 (RS) 已被确立为通过检索叶绿素-a (Chl-a) 等光学活性水质参数来确定内陆水域营养状态的重要信息来源。然而，在全球范围内使用 RS 技术评估内陆水域的营养状态受到检索算法在高度动态和复杂的光学特性上的性能的阻碍，这些光学特性具有许多这些系统的特征。在这项研究中，我们开发了一种新的 RS 方法，以基于中分辨率成像光谱仪 (MODIS) 图像和 Forel-Ule 指数 (FUI) 评估全球内陆水体的营养状态。第一的，FUI 是根据 MODIS 数据通过将天然水的颜色划分为从深蓝色到黄褐色的 21 个指数计算得出的。然后基于原位测量和 MODIS 产品建立 FUI 与营养状态指数 (TSI) 之间的关系。在基于 FUI 的营养状态评估中，采用 645 nm 波段的水离开反射率来区分有色溶解有机物 (CDOM) 主导的系统。在此分析的基础上，开发并应用基于 FUI 的营养状态评估方法，利用 2012 年南方和北方夏季的 MODIS 数据评估分布在世界各地的 2058 个大型内陆水体（表面积 >25 km2）的营养状态。我们的结果表明，可以从 MODIS 中以相当大的准确度（92.5%，R2 = 0. 92) 通过与大范围湖泊的同步原位测量进行比较，基于 FUI 的营养状态评估方法的整体准确度为 80.0% (R2 = 0.75)，经独立数据集验证。在考虑的全球大水体中，贫营养大湖集中在中亚和南美洲南部的高原地区，富营养大湖集中在中非、东亚和北美中北部和东南部。