Precipitation in most part of the Iberian Peninsula is strongly dependent upon the North Atlantic Oscillation (NAO) state. Consequently, a considerable impact of the NAO on precipitation-related variables can be expected. Several previous studies have analyzed the magnitude and spatial extent of such impact, mainly on the winter precipitation field over Iberia. It is worth noting that a large proportion of precipitation over Iberia falls within the winter season. Correlation coefficient values of –0.7 (between the NAO index and precipitation) are typical for the western and southern sectors of the Iberian Peninsula. Thus, winter months with large positive NAO index are usually characterized by lack of precipitation. On the contrary, winter months with large negative NAO index correspond to wetter than usual conditions.

Using NCEP data for the Iberian region we show that NAO influence over two different precipitation-related variables, namely, precipitation rate and precipitable water display different patterns. Precipitable water is shown to be strongly related to the corresponding anomaly fields of temperature while precipitation rate appears to be controlled by the surface vorticity field and associated strength of the tropospheric synoptic activity. In this work, we focus our study on the impact of this important atmospheric circulation mode on the river flow regime of several major Iberian rivers.

In the first part, we analyze the impact of the NAO on the river flow of some of the most important rivers of the Iberian Peninsula, namely the Tagus, the Douro and the Guadiana. We then analyze the subsequent hydroelectric power generation associated with the high and low phases of the NAO index. An empirical relationship between river flow, hydroelectric power output and the NAO state is obtained. Results show that the large inter-annual variability of flow of these three rivers is largely modulated by the NAO. Such modulation, associated with the recent positive trend of the NAO index, might implicate a significant decrease of the available flow. This reduction can represent an important hazard for the Portuguese and Spanish economies due to its negative impact in agricultural yield and hydroelectric power production.

In the second part, we carried out a forecasting experiment. We employ a recently developed ARMA forecasting model of the NAO index to forecast, based on the former empirical model, the river flow and hydroelectric power output. Results show that the modeling has useful one-year-ahead forecasting abilities.