Regional Climate Change in the Western US
This web application is an interactive 2D map that displays in your web browser - no software installation required. To navigate around the map - L click and drag your mouse to pan, use your mouse wheel or the slider control in the upper left to zoom, and Shift L click and drag to define a new map extent. The westwide climate change layers are dynamic time mosaics - look to the lower left of the map and you will see a time slider to select the monthly data.
Natural Resource Managers are increasingly being asked how climate change will affect ecosystem functions and output. Climate scientists have produced numerous models that depict projected changes across large (i.e. continental and global) landscapes. These Global Climate Model (GCM) outputs are often inapplicable to landscapes at the regional and sub-regional scale due to their coarse resolution and uncertaintity at large scales. To address this issue, efforts have been made to downscale these models. Downscaling is a process where more localized climate and weather models are integrated dynamically or statistically into the the GCM's to produce data at a finer resolution. Data from these downscaled models are more relevant at scales typically of concern for land managers.
A consortium of federal agencies (US Forest Service Regions 1 and 6, US Fish and Wildlife Service, and the US Forest Service Rocky Mountain Research Station Boise Aquatic Sciences Lab) contracted with the Climate Impacts Group (CIG) at the University of Washington to produce downscaled models and data for the Columbia, Upper Missouri, Colorado, and Great Basins. California basins added 12/2012.
The CIG selected the 10 best performing GCM models and assembled those into an Ensemble. Ensemble GCM's were selected based on their ability to accurately predict historical summer temperature and percipitation trends. From this suite of 10 models, two models were selected to serve as 'brackets' or high and low extreme values for the downscaling process. These two models - the Parallel Climate Model (PCM1), the Model for Interdisciplinary Research on Climate (MIROC3.2)- were selected based on their delta (change) values from historical conditions for precipitation and temperature during the summer months. The PCM1 model predicts less temperature increase and wetter summers, the MIROC3.2 model greater temperature increase and drier summers, and the Ensemble falling in between these two.
Estimating future greenhouse gas emissions is a key component of a GCM as these compounds in the atmosphere affect the loss and retention of radiant energy at the earth's surface. The IPCC (InterGovernmental Panel on Climate Change) has developed a number of different emission scenarios representing various socioeconomic conditions. The A1B scenario was selected for this study because recent emission patterns suggest that this scenario is more likely to depict mid-century conditions, which are the focus for this assessment.
This GeoBrowser presents temperature and precipitation 6km resolution downscaled data depicting mid-century (2030-2059) and end-century (2060-2099) predicted change from historical (1916-2006) conditions for an Ensemble of the 10 best performing models in the region.
Climate Variable Definitions:
Precip - Precipitation (mm) - Total precipitation in all forms (snow, rain, etc). Average monthly totals.
Temp - Temperature (Celsius) average for each month.
These layers were created to display the magnitude with which the four (PCM1, echam5, hadgem1 and MIROC3.2) models agree in predicted future parameter values and where they diverge. Model agreement layers for precipitation and temperature values were created by calculating the range of the pixel values for each of the four models - difference of the maximum and minmum of the four predicted model values. The lower this value, the more the models tend to agree, higher values indicate increasing divergence of the models. These results are applicable only to these four models and probably grossly underestimate the disagreement among all 20 models considered in the CIG analysis. Additionally, the bracketing models (PCM1 and MIROC3.2) were selected based on their differences in temperature and precipitation in the summer months so disagreement here is probably overemphasized.