Integrated Dataset of Satellite-Derived Water Quality Parameters and Bird Populations in Five Ramsar Wetlands of India Using Sentinel ‑2 Imagery and in ‑Situ Observations

A multi-year dataset of water quality parameters derived from Sentinel-2 imagery was compiled for five Ramsar wetlands in India, three in Kerala (Vembanad, Sasthamkotta, Ashtamudi), one in Punjab (Harike), and one in Tamil Nadu (Point Calimere) using QGIS-defined boundaries and Copernicus satellite data from 2016–2024 (Kerala) and 2018–2024 (others).Sentinel-2 L1C and L2A imagery (Copernicus Open Access Hub), bird population data (eBird, CNHS, Bird Count India), and in-situ water quality field measurements.Sentinel-2 images with ≤11% cloud cover were downloaded. Atmospheric correction was applied to L1C images using DOS1 (QGIS SCP plugin); L2A images were used directly. Thirteen spectral bands were resampled to 10 m resolution and used with regression models based on in-situ measurements to estimate water quality indicators: TSS, turbidity, conductivity, pH, ammoniacal nitrogen, BOD, and DO for Kerala; chlorophyll-a, cyanobacteria, turbidity, CDOM, and DOC for Harike and Point Calimere. Surface reflectance was extracted using Reflectance = DN ÷ 10,000. Data are stored in TSV files by location and date for spatiotemporal analysis, and bird population records from sources like eBird, CNHS, and Bird Count India cover both migratory and resident species across all sites.本数据集针对印度5处拉姆萨尔湿地（Ramsar wetlands）编译得到多时长水质参数数据集，其中3处位于喀拉拉邦（万蓬纳德Vembanad、萨斯塔姆科塔Sasthamkotta、阿什塔穆迪Ashtamudi）、1处位于旁遮普邦（哈里凯Harike）、1处位于泰米尔纳德邦（卡里梅尔角Point Calimere）。数据集以QGIS划定的湿地边界为基础，采用2016–2024年（喀拉拉邦站点）与2018–2024年（其余站点）的哥白尼（Copernicus）卫星影像完成构建。所用配套数据包括哨兵-2（Sentinel-2）L1C与L2A级影像（获取自哥白尼公开获取中心Copernicus Open Access Hub）、鸟类种群数据（eBird、CNHS、印度鸟类计数Bird Count India）以及原位水质野外实测数据。研究筛选云量≤11%的哨兵-2影像进行下载：针对L1C影像，采用DOS1大气校正法（通过QGIS的SCP插件完成校正）；L2A级影像可直接使用。将13个光谱波段重采样至10米分辨率后，结合基于原位实测数据构建的回归模型，可估算得到各类水质指标：喀拉拉邦站点的总悬浮固体（TSS）、浊度、电导率、pH值、氨氮、生化需氧量（BOD）与溶解氧（DO）；哈里凯与卡里梅尔角站点的叶绿素-a、蓝藻、浊度、有色溶解有机物（CDOM）与溶解性有机碳（DOC）。地表反射率的计算方式为：反射率=DN值÷10000。数据按站点与日期以TSV（Tab-Separated Values）文件格式存储，以适配时空分析需求；来自eBird、CNHS、印度鸟类计数等渠道的鸟类种群记录覆盖所有研究站点的迁徙与留居鸟类物种。