Background
The author, who is employed in the Department of Land, Air and Water Resources of the University of California at Davis, writes that the California Central Valley (CV) "is the most agriculturally productive region in the world and home to five million people," and, hence, he says that extreme heat there "is an important concern." However, Grotjahn notes that using a global climate model capable of representing this phenomenon "cannot resolve the complex topography of the region," and that this can only be done "by dynamically downscaling with a high-resolution regional climate model, or RCM," which enterprise thus became the crux of what this paper attempted to accomplish.

What was done
Grotjahn describes the project by stating that it "assesses how well the Community Climate System Model version 4 (CCSM4) simulates the large-scale conditions needed for extreme hot surface temperatures in the CV," which days are associated with the large-scale meteorological pattern (LSMP) that has been described by Grotjahn and Faure (2008). And of this endeavor, Grotjahn writes that "correct simulation of this large-scale pattern is a necessary condition for successful simulation of California extreme hot days by a regional climate model."

What was learned
The UC Davis researcher reports that "CCSM4 can generate a few comparably large circulation index values (implying high surface temperature anomalies) but not as often as in reanalysis data." Also, he says that "the CCSM4 topography does not have a CV, but a broad topographic slope instead," and that although various choices of CCSM4 grid points were tested, "no combination of global model (CCSM4) grid points adequately describes the CV surface temperatures or their association with the upper-air LSMP associated with CV hot spells."

What it means
In commenting on his findings, Grotjahn states that "an RCM is not likely to overcome boundary conditions from a global model that lacks the right large-scale circulation patterns." And in this regard, he notes that "CCSM4 can generate strong enough large-scale patterns for CV hot spells" and that "the patterns can last as long as observed." However, he indicates that "CCSM4 does not create the patterns as often as does the atmosphere." Hence, it would appear that predicting what needs to be known by CV farmers and others in the CV is still pretty much of a hit-and-miss undertaking.

Reference
Grotjahn, R. and Faure, G. 2008. Composite predictor maps of extraordinary weather events in the Sacramento California region. Weather Forecast 23: 313-335.