Ensemble‐Based Spatially Distributed CLM5 Hydrological Parameter Estimation for the Continental United States
Open Access
- 7 February 2025
- journal article
- Published by American Geophysical Union (AGU) in Journal of Advances in Modeling Earth Systems
- Vol. 17 (2)
Abstract
One of the major challenges in large‐domain hydrological modeling efforts lies in the estimation of spatially distributed hydrological parameters while simultaneously accounting for their associated uncertainties. Addressing this challenge is particularly difficult in ungauged locations. With growing societal demands for large‐scale streamflow projections to inform water resource management and long‐term planning, evaluating and constraining hydrological parameter uncertainty is increasingly vital. This study introduces a hybrid regionalization approach to enhance hydrological predictions of the Community Land Model version 5 (CLM5) across the Continental United States (CONUS), with a total of 50,629 1/8° grid cells. This hybrid method combines the strengths of two existing techniques: parameter regionalization and streamflow signature regionalization. It identifies ensemble behavioral parameters for each 1/8° grid cell across the CONUS domain, tailored to three distinct streamflow signatures focused on low flows, high flows, and annual water balance. Evaluating this hybrid method for 464 CAMELS (Catchment Attributes and Meteorology for Large‐sample Studies) basins demonstrates a significant improvement in CLM5 hydrological predictions, even in challenging arid regions. In CONUS applications, the derived spatially distributed parameter sets capture both spatial continuity and variation of parameters, highlighting their heterogeneous nature within specific regions. Overall, this hybrid regionalization approach offers a promising solution to the complex task of improving hydrological modeling over large domains for important hydrological applications.Keywords
Funding Information
- Office of Science
- National Energy Research Scientific Computing Center (BER‐ERCAP0019925)
This publication has 55 references indexed in Scilit:
- Signature‐Domain Calibration of Hydrological Models Using Approximate Bayesian Computation: Empirical Analysis of Fundamental PropertiesWater Resources Research, 2018
- Towards seamless large‐domain parameter estimation for hydrologic modelsWater Resources Research, 2017
- Deriving global parameter estimates for the Noah land surface model using FLUXNET and machine learningJournal of Geophysical Research: Atmospheres, 2016
- Global‐scale regionalization of hydrologic model parametersWater Resources Research, 2016
- Accelerating advances in continental domain hydrologic modelingWater Resources Research, 2015
- Global hydrology 2015: State, trends, and directionsWater Resources Research, 2015
- Optimizing hydrological consistency by incorporating hydrological signatures into model calibration objectivesWater Resources Research, 2015
- Global warming and changes in risk of concurrent climate extremes: Insights from the 2014 California droughtGeophysical Research Letters, 2014
- Are we unnecessarily constraining the agility of complex process-based models?Water Resources Research, 2014
- Global patterns in base flow index and recession based on streamflow observations from 3394 catchmentsWater Resources Research, 2013