• English 
  • Spanish 

Reassessing statistical downscaling techniques for their robust application under climate change conditions

Revista: Journal of Climate
Año: 2013   Volumen: 26
Página inicial: 171   Última página: 188
Estado: Publicado
Archivo PDF: 2013_Gutierrez_JClim.pdf
Enlace al PDF: Online paper
DOI: 10.1175/JCLI-D-11-00687.1

The performance of Statistical Downscaling (SD) techniques is critically re-assessed with respect to their robust applicability in climate change studies. To this aim, in addition to standard accu- racy measures and distributional similarity scores, we estimate the robustness of the methods under warming climate conditions working with anomalous warm historical periods. This validation frame- work is applied to intercompare the performance of twelve different SD methods (from the analogs, weather typing and regression families) for downscaling minimum and maximum temperatures in Spain. First, we perform a calibration of these methods in terms of both geographical domains and predictor sets; the results are highly dependent on the latter, with optimum predictor sets including information of near-surface temperature (in particular 2 meters temperature), which discriminate appropriately cold episodes related to temperature inversion in the lower troposphere.
Although regression methods perform best in terms of correlation, analog and weather generator approaches are more appropriate for reproducing the observed distributions, especially in case of wintertime minimum temperature. However, the latter two families significantly underestimate the temperature anomalies of the warm periods considered in this work. This underestimation is found to be critical when considering the warming signal in the late 21st century as given by a Global Climate Model (the ECHAM5-MPI model). In this case, the different downscaling methods provide warming values with differences in a range of 1 degC, in agreement with the robustness significance values. Therefore, the proposed test for robustness is a promising technique for detecting lack of robustness in statistical downscaling methods for climate change projections.