Climate Change: Assessing Preference Among Surface Water Hydrologic Models for Studies Involving Climate Change
Project ID: 2404
Principal Investigator: Levi Brekke
Research Topic: Water Operation Models and Decision Support Systems
Priority Area Assignments: 2010 (Climate Change and Variability Research)
Funded Fiscal Years: 2009 and 2010
* How and why are models more or less robust (preferable) for simulating runoff in a changed climate relative to that of model calibration?
Natural runoff response to climate change in the Western United States can be simulated using various types of surface water models (e.g., SacSMA/Snow17, VIC, PRMS, Thorthwaite Monthly Water Balance). For Reclamation's longer-term planning evaluations, scoping teams must:
* Decide on which model-type to use given that multiple model-types may have been applied in their study basin
* Demonstrate whether that model-type is preferred among available options.
The proposed research is designed to reveal whether preference exists among surface water hydrologic models as applied to simulate changes in natural runoff under climate change. The approach hinges on designing a calibration and validation scheme involving contrasting historical climates (e.g., calibration on cool-wet years and validation on warm-dry years).
Need and Benefit
Reclamation is capable of using various hydrologic and operations response models for studying climate change effects on Reclamation's water and power operations. Such studies would generally feature an analytical sequence of relating changed climate conditions to changes in surface water supply, water demand, operating constraints, and ultimately operations.
A scoping decision for these types of studies is choosing a surface water hydrologic model "type, " among available options, for assessing natural runoff and water supply response to climate change. There has been a proliferation of model-applications developed for Western United States basins using various types of models (e.g., SacSMA/Snow17, VIC, PRMS, Thorthwaite Monthly Water Balance), which could be construed as being "available." Being an "available" tool has institutional significance for Reclamation's environmental documentation of longer-term planning studies (e.g., National Environmental Policy Act [NEPA], Endangered Species Act [ESA]).
The requirement to use best "available" tools leads to some scoping questions:
* Among available runoff model-types, is there a preferred type for simulating runoff response in a climate that has changed relative to the "model calibration" climate?
* What model-structure features relate to this preference (if it exists)?
Reclamation is currently unable to answer these questions because hydrologic model intercomparison studies have not been conducted with the goal of accurately simulating changes in long-term monthly to annual runoff in a changing climate. There have been model intercomparisons sponsored by National Oceanic and Atmospheric Administration (NOAA) Office of Hydrologic Development (OHD) focused on accurately simulating short-term events at high time/space resolutions.
However, preference under those conditions doesn't imply preference for simulating longer-term runoff under climate change.
There is a related Federal research effort focused on reconciling projected runoff estimates for the Colorado River basin, where multiple model-types are being represented, including SacSMA/Snow17, VIC, and PRMS. However, that effort will conclude by revealing tool-dependent differences in simulated runoff results under common climate change scenarios. It will not reveal tool preference for that application. This is because the tools are not being assessed in a calibration/validation experiment where validation climate (e.g., warm/dry) differs from calibration climate (e.g., cool/wet); and because the calibrations of the model-applications being used were not performed consistently as possible. This project will address both issues. A common model application specification with common calibration objectives will be defined and then uniformly applied in a calibration/validation scheme to all model types, featuring contrasting climates in the calibration/validation periods, respectively. It is expected that the models' different treatment of soil moisture dynamics, land-surface energy balance terms, and snow dynamics would lead to different validation performances, and, thereby, provide insight on which model-structure characteristics may be relatively more important for application in changing climates.
These research findings should provide management benefit by offering scoping guidance for Reclamation NEPA, ESA, and General Planning studies requiring analysis of natural runoff response under climate change. The findings will either indicate that there's a preferred model such application, or that there's no clear preference thereby permitting use of any the model options considered herein. This guidance and associated scoping decision would improve analytical justification within NEPA, ESA, and General Planning documents, thereby limiting risk associated with analytical design challenges in these planning processes. Further, if findings indicate model preference exists and why
Contact the Principal Investigator for information about these documents.