Hot spots of extreme precipitation change under 1.5 and 2 °C global warming scenarios

Climate change hot spots are the regions where the climate variables are particularly responsive to global warming. This study quantifies the hot spots of extreme precipitation based on the simultaneous changes in the intensity (Delta I), frequency (Delta F), and interannual variability (Delta sigma) of extreme precipitation index under the 1.5 and 2 degrees C warming levels with respect to the period of 1986-2015. To investigate whether the hot spots would change as extreme precipitation index alters, precipitation extremes are defined based on the 90th, 95th, 99th, and 99.9th percentiles over the distribution of all wet days during 1986-2005. Daily precipitation from 15 GCMs in the CMIP5 project and the corresponding statistically-downscaled 0.25 degrees x 0.25 degrees dataset (NEX-GDDP) are adopted and compared. A gauge-based global precipitation product is also used to validate the extreme precipitation statistics in both datasets. The result shows that the CMIP5 GCMs and the NEX-GDDP display similar spatial patterns for percentile thresholds, Delta I, Delta F, and Delta sigma at global scale. Then with utilizing the Standard Euclidean Distance method, four aggregated indices (R90, R95, R99, and R99.9) are constructed to represent the comprehensive changes in extreme precipitation indices in the 1.5 and 2 degrees C warming levels. It is found that Sahara emerges as a hot spot of extreme precipitation change in both warming levels for R90, R95, and R99. Model spread in R99.9 is very large indicating large uncertainty in CMIP5 GCMs in the tail of the precipitation distribution. Comparison of the results for the two warming levels indicates that a smaller warming of 0.5 degrees C would lead to a reduction in the magnitude of extreme precipitation by about 7%-8% over global land regions. In addition, the values of Delta I, Delta F, and Delta sigma of P90, P95, and P99 are comparable for both warming levels, while the interannual variability of P99.9 responds more obviously to global warming than the intensity and frequency.