Dataset and source codes for Araki et al., (2026) "Continental-scale prediction of hydrologic signatures and processes"

Abstract

This repository contains code used to generate results for the following manuscript: Araki, R., Holt, A., Hammond, J. C., Husic, A., Coxon, G., and McMillan, H. K. (2026). Continental-scale prediction of hydrologic signatures and processes, Hydrology and Earth System Sciences. Vol 30, 3647–3673, https://doi.org/10.5194/hess-30-3647-2026 This research analyzed streamflow patterns (hydrologic signatures) to predict hydrologic processes across the continental United States. src-signature-prediction.zip: Contains code for the data pipeline, signature analysis, and random forest modeling. Continuously developed at https://github.com/RY4GIT/signature-prediction src-Wetland_GeologicAge_Attributes.zip: Contains code for extracting wetland and geologic age attributes. Continuously developed at https://github.com/RY4GIT/Wetland_GeologicAge_Attributessrc-TOSSH.zip: Contains custom TOSSH toolbox for the signature calculation. Continuously developed at https://github.com/RY4GIT/TOSSH data.zip: Contains the output dataset from the analysis, including observed and predicted signature values. Version note (2026-10-03, version 2.0) This update recalculates four hydrologic signatures so that they use the custom parameter based on regional precipitation patterns and flow levels, as the rest of the analysis. Region-specific recession parameters were previously applied to sig_RecessionAnalysis (outputs RecessionParameters_b) They are now also used for: sig_SeasonalVarRecessions (Recession_a_Seasonality) sig_StorageFromBaseflow (AverageStorage) sig_BaseflowRecessionK (BaseflowRecessionK) Region-specific event-delineation parameters were previously applied to sig_EventGraphThresholds (IE_thresh, IE_thresh_signif, SE_thresh, SE_thresh_signif, avg_IE_SE_thresh, and avg_IE_SE_signif). They are now also used for: sig_EventRR (EventRR)