CR 42:161-176 (2010)  -  DOI: https://doi.org/10.3354/cr00871

ABSTRACT: This paper shows the results of modeling and mapping monthly maximum and minimum temperature, and total precipitation in Egypt with the purpose of obtaining accurate climate maps. A multivariate linear regression model enhanced by spline interpolation was undertaken. Climate variables were obtained from 40 quality controlled and homogenized series for the period 1957 to 2006. The predictors, including geographical variables (e.g. latitude, longitude, altitude and distance to water bodies) and the Moderate Resolution Imaging Spectroradiometer (MODIS) remote sensing indices, were integrated as raster layers in a Geographical Information System (GIS) environment. Inclusion of meaningful remote sensing indices (e.g. the Normalized Difference Vegetation Index and the Normalized Difference Temperature Index) generally improved accuracy of the predictions. The model integrating geographical and remote sensing indices explained an average of 76.5 and 51.7% of the spatial variability of maximum and minimum temperatures, respectively. For precipitation, the model explained an average of 60.2% of the spatial variability during the whole year and 70.7% during the wet season (September to April). The accuracy of the models was assessed through cross-validation between predicted and observed values using a set of statistics including the coefficient of determination (R²), Mean Absolute Error (MAE) and Willmott’s D. The cross-validation results were satisfactory for maximum temperature (average MAE = 1.03°C) and total precipitation (average MAE = 2.73 mm). A poorer fit of the model was obtained for minimum temperature (average MAE = 1.72°C). For each climatic variable, digital maps were finally obtained at a spatial resolution of 1 km. Considering the favourable results obtained using only a small number of observatories, such digital maps have significant potential for the study of climate change and climate impact assessment.

KEY WORDS: Multivariate regression · Temperature · Precipitation · Digital elevation model · GIS · Remote sensing · Egypt