Understanding how Different Versions of Distributed Historical Weather Data Affect Hydrologic Model Calibration and Climate Projection Downscaling

Project ID: 675
Principal Investigator: Subhrendu Gangopadhyay
Research Topic: Water Resource Data Analysis
Priority Area Assignments: 2011 (Climate Change and Variability Research), 2012 (Climate Change and Variability Research)
Funded Fiscal Years: 2011 and 2012
Keywords: None

Research Question

Research questions:

1. How different, in terms of spatial patterns, are the presently available distributed weather datasets in terms of precipitation and temperature?

2. How sensitive is the hydrology model response when calibration is carried out with one of the presently available distributed weather datasets and forced with the other datasets?

3. What are the implications of the climatology differences from the distributed weather datasets for statistical downscaling of general circulation model (GCM) outputs?

Need and Benefit

This proposal is aimed at Reclamation's Safety and Technology (S&T) fiscal year 2011 (FY11) call to respond to the following topics:

2. Obtain Climate Projection Data
* Improved understanding on the strengths and weaknesses of downscaled data and datadevelopment methods (e.g., where and when the stationarity assumption of statistical downscaling techniques may not hold and should be replaced by dynamically downscaled data).

4. Assess Natural Systems Response to Climate Change
* Watershed Hydrology
* Guidance on strengths and weaknesses of available versions of spatially distributed hydrologic weather data that may be used for both watershed hydrologic model development (Topic 4) and in climate model bias-correction (Topic 2).

Water agencies will continue to face decisions about which downscaled data to use and implicitly which downscaling method and spatial resolution to rely upon. As decisions are made, there will be interest in understanding the assessment uncertainties introduced by these choices. However, quantitative information on such uncertainties is limited, setting up the need for evaluations such as this proposed effort. Multiple distributed weather sets are currently available
e.g., the CMIP3 GCM archive was downscaled using the Maurer et al. (2002) distributed weather dataset using a single statistical downscaling algorithm [bias-correction]). The question here is, was Maurer et al. (2002) data the best dataset to be used in this downscaling effort? What would have been the implications of using another distributed weather archive? This proposed effort is aimed to answer these questions. Note that the datasets will be analyzed using two measures: (1) spatial similarity of the distributed precipitation and temperature fields of the study datasets and (2) implications on hydrologic modeling. This would then provide guidance on the choice of datasets for statistical downscaling of GCM outputs used in different types of scale-dependent planning assessments.

Contributing Partners

Contact the Principal Investigator for information about partners.

Research Products

Independent Peer Review

The following documents were reviewed by qualified Bureau of Reclamation employees. The findings were determined to be achieved using valid means.

How Does the Choice of Distributed Meteorological Data Affect Hydrologic Model Calibration and Streamflow Simulations? (final, PDF, 3.5MB)
By Marketa Elsner, Subhrendu Gangopadhyay, Tom Pruitt, Levi Brekke, Naoki Mizukami and Martyn Clark
Publication completed on May 29, 2014

Spatially distributed historical meteorological forcings (temperature and precipitation) are commonly incorporated into modeling efforts for long-term natural resources planning. For water management decisions, it is critical to understand the uncertainty associated with the different choices made in hydrologic impact assessments. This paper evaluates differences among four commonly used historical meteorological datasets and their impacts on streamflow simulations produced using the VIC model.

Last Updated: 9/19/16