Improving Mean Annual Precipitation Prediction Incorporating Elevation and Taking into Account Support Size

Accounting for secondary exhaustive variables (such as elevation) in modelling the spatial distribution of precipitation can improve their estimate accuracy. However, elevation and data are associated with different support sizes it is necessary to define methods combine such data. The paper was aimed compare block ordinary cokriging kriging an external drift estimating annual using covariate. Block used reference a univariate geostatistical approach. In addition, were also taken into account. study area Calabria region (southern Italy), which has spatially variable Mediterranean climate because its high orographic variability. drift, compared cokriging, most accurate approach mean precipitation. three measures accuracy (MAE, absolute error; RMSEP, root-mean-squared error prediction; MRE, relative error) have lowest values (MAE = 112.80 mm; RMSEP 144.89 mm, MRE 0.11), whereas goodness prediction (G) highest value (75.67). results clearly indicated that use always improves estimate, but case areas elevations below 120 m, makes better information estimation than drift. At higher elevations, opposite true: performs cokriging. This takes account size allowed obtain more detailed maps kriging. produced map local details re-evaluation linear regression on estimates.