International Symposium on
Drylands Ecology and Human Security

Go to Arabic site :-)

© 2006 NDRD        Imprint        Disclaimer


Comparative Application of Geostatistical Techniques for Determining Annual Rainfall: A Case Study of Southwestern Iran

Arash Malekian*, Hossein Mohammad Asgari & Gholamreza Zehtabian

University of Tehran, Karaj, Iran


The role of rainfall is essential in arid and semi –arid ecosystems. It is the driving phenomenon of various activities, and in many cases is the only source of life. Measured rainfall data are important to many problems in natural resources management of drylands. For example the ability of obtaining high resolution estimates of spatial variability in rainfall fields becomes important for identification of locally intense storms which could lead to floods. The accurate estimation of the spatial distribution of rainfall requires a very dense network of instruments, which entails large installation and operational costs. It is thus necessary to estimate point rainfall at unrecorded locations from values at surrounding sites.

Geostatistics, which is based on the theory of regionalized variables, is increasingly preferred because it allows one to capitalize on the spatial correlation between neighboring observations to predict attribute values at unsampled locations and does not that ignore both the elevation and rainfall records at surrounding stations.

This study was conducted in southwestern part of Iran. The data set including 20 years of annual average rainfall collected from synoptic stations of the study area were used. First, normality and homogeneity of data were examined. Then variographic analysis using three different techniques including Kriging, Cokriging and Weighted Moving Average (WMA) were applied for predicting annual rainfall. The statistical comparison of the results obtained by three methods showed that Kriging method has the highest accuracy and the mean square error of Kriging prediction is up to half the error produced using other methods. Larger prediction errors are obtained for the two other algorithms (Cokriging and Weighted Moving Average).

Keywords: Rainfall, Interpolation, Geostatistics, Drylands, Iran