Application of Spatial Interpolation Method for Estimating the Spatial Variability of Rainfall in Semiarid New Mexico, USA

Spatial interpolation of rain gauge data is important for ecohydrology study or modelling of land degradation. The monthly rainfall data recorded at 68 stations in Jornada basin was analysed to study the spatial patterns of rainfall. The inverse distance weighting spatial interpolation method was applied to model the spatial variability of rainfall for a wet (1992) and dry (1994) years. The rainfall interpolation was tested for the basin wide region and by constraining the rainfall-interpolation within the study area boundary. The accuracy of the interpolation result was measured by adopting the leave-one-out cross validation method. The result indicates that the rainfall displayed a strong spatial variability trend from the southwest to the northeast. The result from the CV analysis of the total data points for both year showed that the IDW interpolation method produced from data points within the study area boundary produced better fits compared to the CV result from all the data points within the Jornada basin. The result from constraining the rainfall-interpolation within the study-area boundary showed that the interpolation error from the edge effects can be minimised and result of the predicted values showed a closer fit to the measured values.