[general]
name=satellite_derived_bathymetry_for_qgis
qgisMinimumVersion=3.22
version=2
author=Nasef M.Aly
email=Eng.m.nasef2017@gmail.com
description=A comprehensive and automated toolkit for Satellite-Derived Bathymetry (SDB) using machine learning.
about=<h3>Satellite-Derived Bathymetry (SDB) Toolkit</h3><p>This plugin provides a complete, end-to-end ecosystem for Satellite-Derived Bathymetry (SDB) directly within QGIS. It integrates ground truth data preparation with a powerful, automated workflow for image processing, model comparison, and bathymetry generation.</p><p><b>Core Components:</b></p><ul><li><b>ICESat-2 Data Preparation Tool:</b> A utility designed to streamline the use of pre-processed ICESat-2 data (e.g., from SlideRule Earth's CSV output). It allows the user to easily filter the points by confidence level and date, and reproject them to match the project's coordinate system. This prepares a clean, analysis-ready dataset for use in the SDB Master Workflow.</li><li><b>SDB Master Workflow:</b> A fully automated tool that performs preprocessing, runs and compares multiple algorithms, and delivers the best-performing bathymetry map.</li></ul><p><b>Workflow Stages in SDB Master Workflow:</b></p><ol><li><b>Preprocessing:</b> Optional but critical corrections for common errors.<ul><li><b>Sunglint Correction:</b> Implements the <b>Hedley et al. (2009)</b> method to remove sun glint from the water surface.</li><li><b>Water Masking:</b> Automatically isolates water from land using MNDWI/NDWI with an adaptive threshold determined by <b>Otsu's method</b>.</li></ul></li><li><b>Algorithm Comparison:</b> Trains, tests, and objectively compares a suite of scientific models.<ul><li>For complex models (e.g., RandomForest, SVR), it uses <b>Bayesian Optimization with Gaussian Processes</b> for intelligent hyperparameter tuning.</li><li>Includes the classic log-ratio model by <b>Stumpf et al. (2003)</b>.</li></ul></li><li><b>Evaluation & Reporting:</b> All models are validated against unseen test data. A final summary report ranks the models by performance, and the best result is loaded into QGIS.</li></ol><p>This toolkit transforms SDB from a complex, manual task into a standardized, accessible, and scientifically robust process.</p>
homepage=https://github.com/Nasef2017/Satellite-Derived-Bathymetry-for-QGIS
repository=https://github.com/Nasef2017/Satellite-Derived-Bathymetry-for-QGIS
tracker=https://github.com/Nasef2017/Satellite-Derived-Bathymetry-for-QGIS/issues
tags=sdb,bathymetry,icesat,satellite,remote sensing,machine learning,sunglint,hydrography,coastal,water,automation
category=Raster
changelog=Initial release of the SDB Tools suite, featuring the Master Workflow for automated model comparison and prediction.
plugin_dependencies=scikit-learn,rasterio,scikit-optimize,scipy,pandas,numpy,matplotlib