Data and code for "Evaluating the cross-lake transferability limits of machine learning models for Sentinel-3 inland water Chlorophyll-a retrieval"

This archive contains the data and MATLAB code used to reproduce the analyses, model evaluations, tables, and figures for the manuscript: “Evaluating the cross-lake transferability limits of machine learning models for Sentinel-3 inland water Chlorophyll-a retrieval” The study evaluates the transferability of machine-learning Chlorophyll-a retrieval models trained on the Western Basin of Lake Erie and tested on Saginaw Bay of Lake Huron using Sentinel-3 OLCI remote-sensing reflectance and paired in situ water-quality observations. The archive includes raw and processed data, trained MATLAB model files, MATLAB scripts, classification-specific working files, and documentation needed to reproduce the reported model evaluations. The package supports the global/no-classification model and three classification-based adaptive modelling strategies.本存档包含用于复现下述论文的分析、模型评估、表格及图件所需的数据与MATLAB代码： "Evaluating the cross-lake transferability limits of machine learning models for Sentinel-3 inland water Chlorophyll-a retrieval" 本研究采用哨兵-3号（Sentinel-3）OLCI遥感反射率数据及配套原位水质观测数据，对以伊利湖西盆地为训练集、休伦湖萨吉诺湾为测试集的机器学习叶绿素a(Chlorophyll-a)反演模型的跨湖迁移极限进行评估。 本存档包含复现论文所报道的模型评估所需的原始数据、处理后数据、训练完成的MATLAB模型文件、MATLAB脚本、分类专用工作文件及说明文档。该工具包支持全局/无分类模型与三种基于分类的自适应建模策略。