Research during the latest years has indicated a significant connection between climate and solar activity. Specifically, a relationship between Northern Hemisphere air temperature and sunspot cycle length (SCL) has been shown. By using monthly SCL and land air temperature from 1753-1990 (238 years) we show that this relationship also holds for a single observation point in south of Sweden. Using data after 1850 yields a statistically significant linear correlation of 0.54 between SCL and mean temperature. Furthermore, we show that there are indications of a low-dimensional chaotic component in both SCL and the interconnected mean land air temperature. This has important implications for hydrology and water resources applications. By pure definition of chaos this means that it is virtually impossible to make long-term predictions of mean temperature. Similarly, because of the strong connection between temperature and many hydrological components, it is probable that also long-term water balance constituents may follow chaotic trajectories. Long-term projections of water resources availability may therefore be impossible. Repeated short-term predictions may however, still be viable. We exemplify this by showing a technique to predict interpolated mean temperature 6 and 12 months ahead in real time with encouraging results. Improving the technique further may be possible by including information on the SCL attractor. To summarize, research into the possible existence of chaotic components in hydrological processes should be an important task for the next years to come.