Report data are presented in graphical or other visual formats that make them quick to grasp but difficult to analyze numerically or embody in regional studies. In addition, the raw data have undergone substantial filtering and averaging before inclusion. This makes it difficult for researchers to examine regional climate change or incorporate individual station data in other climate-related studies.

It seems natural that the National Climate Data Center (NCDC) would be the source of such data for the United States—and indeed it is—but it’s not readily available in a format that allows for easy import into a GIS. A project at Oregon State University, Parameter-elevation Regressions on Independent Slopes Model (PRISM), is the primary source of raster climate data for the United States, providing monthly averaged daily temperature minimum and maximum as well as precipitation totals in a format that can be imported into a GIS. These rasters comprise interpolated data, but their nature doesn’t allow flexibility in the length of the averaging period or access to the individual station data used in the interpolation.

A tool developed at California State University, Northridge, now converts NCDC historical climate data to a feature-class format. A menu-driven interface provides access to point temperature and precipitation data that can be interpolated easily through the tools provided in the Spatial Analyst toolbox to generate raster layers for any region in the United States averaged over any desired temporal period. Thirty-year baseline averages for the 1950-1980 period, for example, have been pre-calculated and stored in a geodatabase to permit easy calculation of anomalies relative to this period, providing the framework for climate-change studies.

This GIS climate tool was written in Visual Basic using ArcObjects and reads data from the daily weather observations of more than 20,000 cooperative weather stations in the United States. Although the period of record can begin as early as 1850, most station histories begin in 1948 or soon thereafter. The data were obtained from NCDC (www.ncdc.noaa.gov/oa/climate/climatedata.html).  

Figure 1 shows the locations of the weather stations within the western United States. The user interface allows desired stations to be selected from a dropdown menu or via feature selection within the map. A second menu allows users to select the time period of interest: a specific date, month, season or year, or the average of any one of these over a range of user-specified years.

Figure 2. A menu-driven interface provides access to point temperature and precipitation data to generate raster layers for any region in the United States.

Computed data, including means and standard deviations, then are saved to a feature class with the option to map any of the weather elements (see Figure 2). Plotting options include the anomaly with respect to the long-term (30-year) mean and standard error in addition to the mean and standard deviation of the selected variable.

One study explores the spatial extent of the effects of El Niño and La Niña within the western United States. Public perception of these, often based on the popular media’s presentation, is that El Niño causes wet winters in California and La Niña causes dry winters, but the effects of these weather systems can extend beyond this and actually reverse when moving northward along the west coast from California into Oregon and Washington.

Figure 3 shows the 30-year mean (1950-1979) seasonal precipitation for the winter (December, January, February) season. Using the GIS climate tool, a student can map the anomalies in winter precipitation for an El Niño or La Niña year relative to the long-term average via simple menu selection (see Figure 4).

These maps illustrate the complementary nature of the two effects. El Niño effects are most pronounced in bringing extra rainfall to the coast of California, but tend to bring drier weather to most of Oregon and Washington. The effects of La Niña also are most severe along the coast, causing drought in much of California, but bringing heavy rain to the coasts of Oregon and Washington. Climate change is likely to cause changes in the frequency, intensity and regional effects of these two systems.

Other lessons in regional climate and climate change can be learned from examining the diurnal temperature range. This is the difference between the daily temperature maximum and the night-time temperature minimum, and it’s strongly influenced by cloud coverage, proximity to the ocean and topography. The seasonal variation in this metric is evident in Figure 5, which shows the wintertime (top) and summertime (bottom) diurnal temperature range for the western United States. This range is much greater in summer than winter, particularly in the inland areas, due to greater insolation in the daytime in summer and lack of cloud cover over inland areas, particularly in summer.

The diurnal temperature range has been a subject of some scrutiny by climate scientists in recent years. It appears as if the signal for global warming is stronger in night-time minimum temperatures than daytime maxima. This is likely the result of increased cloud cover together with the urban "heat-island" effect.

One feedback effect of global warming is an increase in the concentration of water vapor in the atmosphere, which can lead to an increase in cloudiness. Over a 24-hour period, low clouds have a net cooling effect, whereas high clouds have a net warming effect through the absorption and emission of thermal infrared radiation. However, all clouds have a heating effect at night, when there’s no sunlight, so an increase in cloud cover would cause night-time temperatures to increase.

To study this, the aforementioned climate tool can be used to examine changes in temperature. Through menu selection, users can map the difference between the temperature of the last few decades and that of the 30-year mean stored in the geodatabase.

Figure 6 shows this for the period 1975-1999, illustrating the night-time minimum temperature anomaly (top) as well as the daytime maximum (bottom). Throughout nearly all of Arizona, California, Nevada, Oregon and Washington, the minimum temperature has increased, whereas the maximum temperature shows a much more mixed result.

The chaotic nature of these climate-change patterns often is overlooked in the type of globally averaged data so often quoted. Global-warming trends are computed by statistically averaging data from thousands of sources after they have been through extensive quality control and filtering. These globally averaged statistics can be misleading; there’s significant variation in weather and climate on a local and regional basis. When individual station data are employed, climate change results can be much more mixed.

Author’s Note: The California State University, Northridge, GIS climate tool is currently undergoing beta testing by students in its geography department, and it’s being used to provide climate data as needed in research projects. It’s being used to investigate regional patterns of climate change, urban heat-island effects and the regional influence of the El Niño Southern Oscillation. It’s also being used to provide researchers with historical data in a format compatible with GIS studies. Plans are for the tool to be made available online through a GIS interface.