Tuesday, November 3, 2009
Convention Center, Exhibit Hall BC, Second Floor
Understanding the spatial variability of precipitation is important for agriculture and natural resource management as well as many other disciplines. Its variability includes the amount of water, spatial and temporal distribution as well as the sequence of wet and dry days. Several statistical methods have been used to estimate and describe both the spatial and temporal variability of rainfall. However, there are a few studies that have been conducted for daily precipitation. The goal of this study was to compare the estimation of the spatial variability of daily rainfall using kriging and co-kriging approaches. Daily weather data from 70 weather stations that are part of the Georgia Automated Environmental Monitoring Network were used for geostatistical analysis. Unidirectional isotropic variograms and cross variograms were fitted to Gaussian, spherical, exponential and linear models. Co-kriggin, involved the elevations of 221 locations as well as their coordinates. For each day of the year, separate maps for estimated total rainfall and its standard error based on either kriging or co-kriging were developed. Cross validation for both kriging and co-kriging were also conducted using simple linear regression. There was a significant positive relationship among the number of stations that reported rain and the estimated parameters of the semivariogram models fitted. The relationship between daily total rain and elevation varied; some days were positive and other days were negative. However, not all of them were statistically significant. Both the kriging and co-kriging methods produced bias. However, the relative bias for co-kriging was less than for kriging and it also produced a better estimate for the non-sampled points compared to kriging. The implementation of a daily precipitation estimation system based on a co-kriging approach is a very useful tool in agriculture because is one of the most important inputs for the use and management of water and crop.