Evaluación de datos interpolados para la detección de índices de extremos térmicos

Abstract

In this study we perform a comparison of extreme indices from observed and interpolated dataset. In evaluation studies on simulated data is required to have grid datasets. However, observed data are irregularly distributed and the interpolation procedure may cause some smoothing on datasets mainly in the extreme indices. Therefore, it is interesting to learn about the effect of interpolation on extreme indices to assess their applicability for impact studies. Spatial and temporal analyses were applied to daily maximum and minimum temperatures and also to the derived extremes, summer and frost days. The dataset used are E-OBS from the ENSEMBLES European Project and observations from the Spanish Meteorological Agency (AEMeT), completed with European Climate Assessment & Dataset (ECA&D) observations from Portugal. Before making the comparison, a control analysis was applied. The results are consistent with the increase greenhouse gases causing more summer days and fewer frost days. The trend and inter-annual variability show good agreement between both dataset. Nevertheless, biases of the mean values are obtained in higher elevation places. These results have to be considered in future climate extreme projections.