Spatial Data Analysis

Spatial Data Analysis of California Climate:


The analysis takes the average temperatures recorded at each climate station in California over the years from 1961 to 2008 and analyze the covariograms and deviations of temperature comparing to the average temperature from 1961 to 1990. The project makes use of the R package “maps” and “fields”, which help map the state (in this case, California), and graph the temperature changes throughout the state using a color gradient. By looking at the deviations of temperature from 2001 to 2008 on eight separate graphs, we were able to see if there are major changes in temperatures in some area, and speculate what may be some of the reasons that cause such changes.


We first take the Data and plot the average temperatures from 1961 to 1990 at each station location. One of the main functions that made the analysis possible was the way to calculate covariogram. Covariogram is calculated assuming stationarity and isotropy hold for the inputted data, which are the temperature recorded and the location of the station (Graph 1). Covariance is calculated using the variances of stations throughout the years (Graph 3). The function then estimates by using the relationship between covariance and distance, which is a matrix containing each station’s distance from every other stations, and output the estimated temperature of each longitude/latitude location of California using a nonlinear regression line (Graph 2). Using such method, we are able to look at California’s temperature means and deviations as a whole, instead of just spots of locations.

Graph 4 composed of 8 graphs, showing the deviation of grid in comparing to the average temperature respective to each location obtained from the data of Table 1. The deviation is shown on a color scale from red to blue. With red being positive deviation and blue being negative. The scale goes from +18 to -18, center at 0, which is white. From the data, the results showed that the largest positive deviation happened in 2008, with deviation being a little bit over 17. Lowest deviation occurred in 2007, being a little bit lower than -12.


The overall trend seems to be that each year, almost all California looks to be a little bit hotter comparing to its historical average temperature. This can be explained by the recent Global Warning, with locations being about 2-3 deviations higher than average. Deviation reaches its peak in 2008 in San Bernardino area. Such climate change can explained due to San Bernardino’s dry conditions combined with hot temperatures and gusty winds generated by the several major wildfires. In 2007, Central California, around Toulumne, Mariposa, and Mono counties, had a high deviation, which is possibly caused by the high pressure over the area that year, which resulted in hot conditions. Deviation was the lowest in 2007 in Central California, around Inyo and Tulare counties. Large amount of thunderstorms caused many problem of power outrage in Central California that year, and such large rainfall in the area that year may be the cause of the fall in temperature.

In an overall trend, California overall seems to be hotter in 2001 to 2004 then 2005 to 2008. The general outlook of recent years appears whiter – meaning much less deviations from historical averages. This may be due to the effort of stopping Global Warming, which leads to a more stabilize temperature.

%d bloggers like this: