Untitled DocumentDownscaling of climate projections for the U.S. Southwest
Macro Theme Area:
Integrated Modeling [Project ID: M35]
PI:
Francina Dominguez
CO-PI(s):
Julio Canon, Juan Valdes
Basin focus:
Regional SW
Specific area in
basin /
field sites:
N/A
Summary/Goals: Our goal is to evaluate the IPCC Gobal Climate Model's (GCMs)representation of historical and future climate in the Southwest and generate high-resolution precipitation and temperature projections for the Southwest based on IPCC scenarios. The coarse-resolution projections generated by GCMs participating in the IPCC Fourth Assessment Report, are downscaled to an appropriate resolution for hydrologic models and impact assessment studies. In addition, future ENSO projections are incorporated into our downscaling estimates.
Our primary motivation for this work is the fact that many hydrologic impact assessment studies rely on a downscaling technique developed by Wood et al., 2004. This technique is used in the recent data archive containing fine-scale climate projections funded in part by the Bureau of Reclamation and the U.S. Department of Energy's Office of Fossil Energy through the National Energy Technology Laboratory. Wood et al.'s methodology is very attractive because it combines bias correction with spatial downscaling in a very computationally efficient way. However, one of the largest drawbacks of the method is the use of climatological values to perform the spatial downscaling. This means that the spatial distribution of climate variables is not dynamic in time. Furthermore, the spatial resolution of the available data archive is 1/8° (12.5km).
Activities and outcomes during past year:
Contribution to Science: Our work shows that the projected future aridity of the region will be dramatically amplified during La Nina conditions, as anomalies over a
drier mean state, and will be characterized by higher temperatures and lower precipitation than the
projected trends. These results have important implications for water managers in the Southwest who must prepare for more intense winter aridity associated to future ENSO conditions. In addition, our statistical downscaling technique uses spatial distributions of precipitation and temperature that are variable in time with ENSO variability being the principal driver. We use an innovative statistical technique called Multichannel Singular Spectrum Analysis (M-SSA) to perform the spatial downscaling. Our downscaling effort uses PRISM data at a 4km resolution. Because of the computational effort of these estimates, two models are used: the HAD-CM3 (UK) and the MPI-ECHAM5 (Germany), these two models have been shown to provide the best climate estimates over the Southwest.
1-The downscaled data at a 4km resolution is now available for researchers through the SAHRA website.
2-One article: Dominguez, F. and J. Caņon and J. Valdes, 2008 IPCC-AR4 climate simulations for the Southwestern US: the importance of future ENSO projections. has been accepted with minor revisions in the Journal Climatic Change.
3-One article: J. Caņon, F. Dominguez and J. Valdes, 2008 Downscaling climate variability associated with quasi-periodic climate signals: a new statistical approach is pending revisions in the Journal Climatic Change.
4-The Dynamical Downscaling using the WRF model has begun. The seasonal forecast is being used first.
Plans for the upcoming year:
Our activities include :
The dynamical downscaling using WRF will continue throughout the year. After the seasonal downscaling produces realistic results we will move on to the IPCC downscaling.
