Identification of external influences on temperatures in California

We use nine different observational datasets to estimate California-average temperature trends during the periods 1950-1999 and 1915-2000. Observed results are compared to trends from a suite of climate model simulations of natural internal climate variability. On the longer (86-year) timescale, increases in annual-mean surface temperature in all observational datasets are consistently distinguishable from climate noise. On the shorter (50-year) timescale, results are sensitive to the choice of observational dataset. For both timescales, the most robust results are large positive trends in mean and maximum daily temperatures in late winter/early spring, as well as increases in minimum daily temperatures from January to September. These trends are inconsistent with model-based estimates of natural internal climate variability, and thus require one or more external forcing agents to be explained. Observational datasets with adjustments for urbanization effects do not yield markedly different results from unadjusted data. Our findings suggest that the warming of Californian winters over the twentieth century is associated with human-induced changes in large-scale atmospheric circulation. We hypothesize that the lack of a detectable increase in summertime maximum temperature arises from a cooling associated with large-scale irrigation. This cooling may have, until now, counteracted summertime warming induced by increasing greenhouse gases effects.