Processed Output Data and Reproducibility Scripts for: "From Snowmelt-Buffered to Rain-Dominated: Projecting Non-linear Hydrological Extremes via Physically-Informed Deep Learning"

This repository contains the processed model outputs, aggregated forcing data, and analytical scripts necessary to reproduce the research findings presented in the manuscript, From Snowmelt-Buffered to Rain-Dominated: Projecting Non-linear Hydrological Extremes via Physically-Informed Deep Learning.

This archive provides the structured, processed datasets (in CSV format) derived from our physically-informed deep learning framework. Due to the substantial storage footprint of the raw CMIP6 forcing data, this repository focuses exclusively on the aggregated time-series outputs that are directly utilized for projecting extreme events. The datasets are systematically organized according to the experimental design and Shared Socioeconomic Pathways (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5).