Climate change projections for winter streamflow in Douro river

Abstract

[EN]Climate change projections for the winter streamflow of the Douro River have been obtained for the period 2071-2099, using the Principal Component Regression (PCR) method. The winter streamflow time series (January to March) from eight stations distributed over the basin, covering the period 1950-2011, were used as predictand variables, while the principal components (PCs) of the winter (December to February) anomalies of sea level pressure (SLP) were used as predictors of the streamflow for the development of a statistical downscaling model. The period 1950- 1995 was used for the calibration of the regression model, while 1996-2011 was used as validation period. The statistical downscaling model fitted from the observational SLP data was applied to the SLP outputs of three GCMs for the period 2071-2099, under the climate change scenarios RCP2.6, RCP4.5 and RCP8.5. The main result obtained is that all models and scenarios project a generalized decrease in the winter streamflow of the Douro River.

[ES]Se han obtenido proyecciones de cambio climático para el caudal de invierno del Río Duero, para el periodo 2071-2099, usando el método de Regresión por Componentes Principales (PCR). Las series temporales de caudal (de enero a marzo) de ocho estaciones distribuidas a lo largo de la cuenca, cubriendo el periodo 1950-2011, han sido utilizadas como variables predictando, mientras que las componentes principales (PCs) de las anomalías de la presión a nivel del mar (SLP) en invierno (diciembre a febrero) fueron usadas como predictores del caudal para el desarrollo de un modelo de downscaling estadístico.