Selection of an Appropriate Interpolation Method for Rainfall Data In Central Nigeria

Abstract

Interpolation technique can be used to establish the rainfall data at the location of interest from available data. There are many interpolation methods in use with various limitations and likelihood of errors. This study applied five interpolation methods to existing rainfall data in central Nigeria to determine the most appropriate method that returned the best prediction of rainfall at an ungauged site. The methods include the inverse distance wieight method and four variants of universal kriging (spherical,exponential,gaussian and power). The Guassian model of Kriging yielded the least root mean square error for monthly rainfall interpolation and is therefore recommended for use on monthly rainfall data in Nigeria.

Key Words: Rainfall, Boxplots, Kriging, Interpolation,Variograms, Gaussian