Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

The Influence of Different Methods of Interpolating Spatial Meteorological Data on Calculated Irrigation Requirements

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  Applied Engineering in Agriculture. 27(6): 979-989. (doi: 10.13031/2013.40625) @2011
Authors:   J. Rolim, J. Catalão, J. Teixeira
Keywords:   Spatial interpolation, Agro-meteorological variables, Irrigation, Nearest neighbor, Inverse distance weighting, Least squares collocation

Irrigation simulation models have become increasingly accurate in the estimation of irrigation requirements. Accurate input data for these models are needed to take full advantage of the increased accuracy. Climatic input data are often supplied by networks of meteorological stations that provide spatially distributed data that need to be interpolated for a given site. The main objective of this study was to evaluate the influence of the different methods of interpolation of the climate data on the estimated irrigation water requirements. Software was developed to perform the spatial interpolation of the meteorological data series recorded by a network of weather stations to the center of each plot. The interpolation methods studied were the nearest neighbor, inverse distance weighting, and least squares collocation methods. Accuracy assessment and comparative analyses were performed for a network of weather stations in southern Portugal for evapotranspiration, wind speed, relative humidity, and precipitation. The impact of the interpolated meteorological data on the accuracy of the calculated crop water requirements was determined from a soil water balance model (IrrigRotation). The accuracy assessment for evapotranspiration suggested that the relative error ranged between 10% and 15%. For precipitation, the relative error ranged between 44% and 60%, showing higher spatial variability and greater difficulty in interpolating this variable. The least squares collocation and inverse distance weighting methods yielded only a slight improvement in the accuracy of the interpolated meteorological data when compared with the most commonly used method, Thiessen polygons (the nearest neighbor method). For the irrigation requirements, the value of the relative error was, on average, 16%, 17%, and 14% for to the nearest neighbor, inverse distance weighting, and least squares collocation, respectively, which correspond to deviations in the irrigation requirements of 79, 86, and 67 mm, respectively. Depending on the irrigation method, this average deviation can represent the saving of some irrigation events. The maximum relative error value for the irrigation requirements was 45% (165 mm), and the minimum relative error was 5% (27 mm). Thus, the spatial variability of meteorological variables has a considerable impact on the accuracy of the calculation of irrigation requirements, with implications for the amount of water used in irrigation. The comparison between the least squares collocation and the inverse distance weighting methods showed identical interpolation performances; however, because the inverse distance weighting is a much simpler method, it can be recommended from the practical point of view.

(Download PDF)    (Export to EndNotes)