Evaluation and Estimation of the 24 hour Maximum Rainfall in the Provice of Golestan

Authors

Abstract

Spatail variability of rainfall has diverse effects on the management and planning of water resources on a watershed, as well as on the national scale. Spatial variability analysis of different parameters, e.g. rainfall, can be very important in estimating the water balance uncertainties and in the decision-making processes. In the presence of Spatail correlation between observations, application of Geostatistical interpolation methods provides usually provides close estimation of the sought values. This technique was implemented in the deformination of the abundance of the 24 hr maximum rainfall of 27 rain guaging stations in Golestan plain. The methods used were ordinary kriging, cokriging and weighted moving average. Cross validation was used to evaluate different interpolation methods. Statistical analysis showed that winter rainfall had the smallest coefficient of variation in comparison with the other time periods used. Geostatistical analysis showed a strong spatial correlation of the data; the best semivariogram model was spherical. The cross-validation results indicated that except for winter, cokriging results in a smaller error for rainfall estimation. Moreover, the generated rainfall maps indicate that the northern region of the presence reccives leave amount of precipitation.