Data and code for: Transformative recovery in an ethnic-minority mountain watershed — a cost-of-success framework for restoration governance (Reshuihe, Liangshan, China)

This repository contains the processed data and analysis code supporting the studyof state-led ecological restoration in the Reshuihe watershed, Liangshan Yi AutonomousPrefecture, Sichuan, China. Contents: 01_remote_sensing/ - Reshuihe_DEM_30m.tif, Reshuihe_Hillshade_30m.tif — Copernicus GLO-30 terrain - Reshuihe_NDVI_2015/2020/2025.tif — Landsat-derived NDVI composites - Reshuihe_NDVI_SenSlope.tif, Reshuihe_MK_Pvalue.tif — pixel-wise Mann-Kendall trend analysis (2000-2025) - Reshuihe_BurnFrequency.tif — multi-decadal burn frequency - Reshuihe_LandCover_WC2021.tif — ESA WorldCover 2021 - Reshuihe_watershed_boundary.shp (+ .shx/.dbf/.prj/.cpg) — watershed boundary - Reshuihe_AnnualBurnArea / MonthlyBurnArea CSV — burned-area time series 02_survey_sem/ - Reshuihe_SEM_data_clean.csv — de-identified household survey data (n = 791) used for single-group and multi-group structural equation modelling - sem_direct_effects.csv, sem_indirect_effects.csv, sem_r_squared.csv — SEM output 03_code/ - GEE_annual_NDVI_export.js — Google Earth Engine script for annual NDVI export - compute_burnrate_by_landcover.py — reproduces burn rate by land-cover class - make_hires_figs.py — figure assembly 04_figures/ - Figures 1-6 as published (600 dpi PNG) Raw interview transcripts are not included owing to confidentiality agreements withinformants and Institutional Review Board protocols protecting minority-communityparticipants.