Abstract. In coastal waters, colored dissolved organic matter (CDOM) is often present in moderate to high concentrations due to input from nearby landmasses and rivers. CDOM absorbs light sharply in the UV-blue region, causing waters to appear yellow or brown. This has a significant effect on optical and biogeochemical processes in coastal waters, as it inhibits the penetration of light within the water. It is thus important to be able to monitor CDOM levels on a regular basis. A model is presented linking Sentinel-2 reflectance data to the CDOM absorption coefficient. Such a model is useful for monitoring the CDOM absorption in coastal waters due to the frequency of Sentinel-2 passes and relatively high-spatial resolution. This model was developed and validated using concurrent field measurements of above-surface remote-sensing reflectance and the in situ CDOM absorption coefficient, from the coastal waters of Singapore. The leave-one-out cross-validation scheme was used to validate the model due to the small number of data points. The model is further tested with Sentinel-2 images of the same locations, not more than one day apart from the in situ measurements.

中文翻译：

---

使用新加坡沿海水域的 Sentinel-2 数据估计有色溶解有机物

摘要。在沿海水域，由于来自附近陆地和河流的输入，有色溶解有机物 (CDOM) 通常以中等至高浓度存在。CDOM 强烈吸收紫外线蓝色区域的光，导致水呈现黄色或棕色。这对沿海水域的光学和生物地球化学过程具有显着影响，因为它抑制了光在水中的渗透。因此，能够定期监控 CDOM 级别非常重要。提出了一个模型，将 Sentinel-2 反射率数据与 CDOM 吸收系数联系起来。由于 Sentinel-2 通过的频率和相对较高的空间分辨率，这种模型可用于监测沿海水域的 CDOM 吸收。该模型是使用来自新加坡沿海水域的地表遥感反射率和原位 CDOM 吸收系数的同步现场测量来开发和验证的。由于数据点数量少，采用留一法交叉验证方案来验证模型。该模型使用相同位置的 Sentinel-2 图像进行进一步测试，与原位测量相隔不超过一天。