SI of Regional Flood Frequency Analysis from Paleoflood Data Using Machine Learning Donor Selection

Regional Flood Frequency Analysis from paleoflood data using Machine Learning donor selection This repository contains the main Python and R code implementing the methods used in our manuscript titled "Regional Flood Frequency Analysis from paleoflood data using Machine Learning donor selection", currently under review at Water Resources Research (WRR). The code covers the main stages of the proposed framework: selection of environmental descriptors, dimensionality reduction of the environmental space, pooling-group (donor-site) formation using both classical and non-systematic-information-based approaches, and regional regression of standardization parameters against basin area. It also provides the scripts and data needed to reproduce the figures, as well as images analogous to those in Supporting Information S7 for each individual site–PaCMAP-scenario combination. More information in README.txt