Time-scale and extent at which large-scale circulation modes determine the wind and solar potential in the Iberian Peninsula

• Published in: Environmental Research Letters
• Main Authors: Sonia Jerez, Ricardo M Trigo
• Format: Article in Journal/Newspaper
• Language: English
• Published: IOP Publishing 2013
• Subjects: North Atlantic Oscillation; East Atlantic; Scandinavian pattern; wind power; solar power; Iberian Peninsula; Environmental technology. Sanitary engineering; TD1-1066; Environmental sciences; GE1-350; Science; Q; Physics; QC1-999; North Atlantic; North Atlantic oscillation
• Online Access: https://doi.org/10.1088/1748-9326/8/4/044035; https://doaj.org/article/d93c57509eaa436b86ab5c793c1ea53c

The North Atlantic Oscillation (NAO), the East Atlantic (EA) and the Scandinavian (SCAND) modes are the three main large-scale circulation patterns driving the climate variability of the Iberian Peninsula. This study assesses their influence in terms of solar (photovoltaic) and wind power generation potential (SP and WP) and evaluates their skill as predictors. For that we use a hindcast regional climate simulation to retrieve the primary meteorological variables involved, surface solar radiation and wind speed. First we identify that the maximum influence of the various modes occurs on the interannual variations of the monthly mean SP and WP series, being generally more relevant in winter. Second we find that in this time-scale and season, SP (WP) varies up to 30% (40%) with respect to the mean climatology between years with opposite phases of the modes, although the strength and the spatial distribution of the signals differ from one month to another. Last, the skill of a multi-linear regression model (MLRM), built using the NAO, EA and SCAND indices, to reconstruct the original wintertime monthly series of SP and WP was investigated. The reconstructed series (when the MLRM is calibrated for each month individually) correlate with the original ones up to 0.8 at the interannual time-scale. Besides, when the modeled series for each individual month are merged to construct an October-to-March monthly series, and after removing the annual cycle in order to account for monthly anomalies, these correlate 0.65 (0.55) with the original SP (WP) series in average. These values remain fairly stable when the calibration and reconstruction periods differ, thus supporting up to a point the predictive potential of the method at the time-scale assessed here.