CTSM input datasets for simulating prescribed and cultural burning effects on low flows in the Salmon River watershed, Klamath River Basin, California

Overview This deposit contains the five input files required to reproduce the Community Terrestrial Systems Model (CTSM) simulations reported in Cheng et al. (2026), which evaluate how Traditional Ecological Knowledge (TEK)-informed prescribed and cultural burning affects low-flow dynamics and salmonid habitat in the Salmon River watershed of the Klamath River Basin, northern California. The burning targets encoded in these files were co-developed with cultural practitioners and Department of Natural Resources staff of the Karuk Tribe through an iterative co-design process described in the associated paper. They represent aspirational land cover trajectories grounded in Karuk management priorities, not forecasts of expected land cover change. All simulations were run with CTSM in Satellite Phenology (SP) mode at 1-km spatial resolution. Files are in netCDF CDF-5 format (cdf5 in the file names); readers require a netCDF library built with CDF-5 support. File inventory domain.klamath.ctsm.cropped.230717.cdf5.nc — Domain file defining the 1-km model grid cropped to the study region: grid cell center and vertex coordinates, land fraction, land/ocean mask, and cell area. Shared by all simulations. klamath.optz_params.cdf5.nc — Calibrated CTSM parameter file. Default parameters produced substantial streamflow biases against USGS gage 11522500 (Salmon River); parameters were optimized using Adaptive Surrogate Modeling-based Optimization (ASMO), improving daily Kling-Gupta Efficiency from 0.49 to 0.62. This file supersedes the default CTSM parameter file for all simulations reported in the paper. surfdata_klamath_hist_78pfts_CMIP6_simyr2005_c220513_OM_HRES.NAT_PCT.cdf5.nc — Baseline (historical) surface dataset. Derived from the standard CMIP6 78-PFT surface dataset for simulation year 2005, regridded to the 1-km domain. Natural vegetation is dominated by needleleaf evergreen trees (96.3% of vegetated cover), reflecting a century of industrial timber management. Used in the Baseline experiment (1981–2020) and in all experiments holding land cover at historical conditions. surfdata_..._NAT_PCT.2050.cdf5.nc — Mid-21st century (2050s) burning-scenario surface dataset. Represents a static land cover snapshot under a near-term, planning-constrained burning trajectory built from existing fire management instruments (Western Klamath Restoration Partnership planning areas; Community Wildfire Protection Plans for Happy Camp, Salmon River, Somes Bar, and Orleans). surfdata_..._NAT_PCT.2100.cdf5.nc — Late-21st century (2100s) burning-scenario surface dataset. Represents basin-wide application of the full TEK-informed spatial zonation (riparian, oak woodland, serpentine meadow, and high-elevation conifer–meadow mosaics organized by distance from river corridors and elevation). The two scenario surface datasets were generated by modifying the natural-vegetation PFT percentage field (and, by construction, bare-ground fraction) of the baseline file; all other surface properties — soil texture, organic matter content, topographic parameters, and non-vegetated land unit fractions — are inherited unchanged. This isolates the land cover signal in the factorial attribution experiments.