This repository contains codes for a study titled "Evaluation of CMIP6 HighResMIP for hydrologic modeling" submitted to Water Resources Research (December, 2022) by Alexander T. Michalek, Gabriele Villarini, Taereem Kim, Felipe Quintero, Witold F. Krajewski, and Enrico Scoccimarro.

The resources include R codes for data analysis, figures, and precipitation bias-correction and downscaling. Additionally, codes are provided related to the setup of the hydrologic model (HLM) utilized in the study and found at https://asynch.readthedocs.io/en/latest/index.html. Finally, a subset of data from the simulations is provided for which the analysis is conducted. Full simulation datasets and CMIP6 forcings are not provided as they are too large to store and can be provided upon reasonable request.

Abstract:
The High-Resolution Model Intercomparison Project (HighResMIP) experiments from the Coupled Model Intercomparison Phase 6 (CMIP6) represent a broad effort to improve the resolution and performance of climate models. The HighResMIP suite provides high spatial resolution (i.e., 25- and 50-km) forcings that have been shown to improve the representation of climate processes (i.e., rainfall, tropical cyclones, atmospheric rivers). However, little is known about the suitability of the HighResMIP suite for hydrologic applications. Here we utilize forcings from the HighResMIP models to simulate annual maximum discharge with the Hillslope-Link Model (HLM) at 1000 communities across Iowa. First, we assess whether the runoff from the climate models can be directly routed through the river network model in HLM to estimate annual maximum peak discharge. We find that the runoff forcing can simulate annual maximum discharge magnitude similar to modeled observations for the historic period. Next, we force the HLM with precipitation, temperature, and evaporation from the HighResMIP models to simulate flood peaks, finding that they also translate well hydrologically. To further improve the performance of these models and their suitability for community-level assessment, we use nine different statistical approaches to bias-correct and downscale HighResMIP precipitation to a 4-km resolution. We find that bias-correction and downscaling of climate model precipitation improve hydrologic modeling performance. Finally, annual maximum discharge from precipitation forcings show a projected increase across much of Iowa, stronger than what is obtained using runoff-driven projections.