Projected signals in climate extremes over China associated with a 2 °C global warming under two RCP scenarios

ABSTRACT Using daily output from 29 climate models provided by the Coupled Model Intercomparison Project Phase 5, we project signals in 12 extreme temperature indices and 12 extreme precipitation indices relative to 1986–2005 over China associated with a 2 °C global warming above pre‐industrial levels under representative concentration pathways 4.5 (RCP4.5) and 8.5 (RCP8.5). The model output reflects the following projected changes: (1) It is robust and statistically significant that warm extremes are more frequent, more persistent, and more intense than those during the baseline period of 1986–2005, and local signals emerge from natural internal variability over most of China. In particular, southern China faces severe heat stress in summer based on warm extreme indices. (2) It is robust and statistically significant that there are fewer cold extremes in China. Most models show no significant changes in the longest duration and intensity of most cold extremes in southern China and northwest China. (3) The multi‐model median shows that more frequent and more intense wet extremes that deliver greater amounts of extreme precipitation occur in China, with a regional mean increase of 16.7–42.8 mm (8–42%) in total amount of annual wet extremes, but these changes are not significant and do not exceed natural internal variability over most of the country. Spatially, the Tibetan Plateau and northeast China have robust and significant changes in part of precipitation extremes, and some changes begin to emerge from natural internal variability at local scale. There are benefits to limiting global warming for China, including less frequent and less persistent warm extremes when comparing 1.5 °C with 2 °C of global warming and a later occurrence of significant changes in climate extremes when compared an intermediate mitigation scenario RCP4.5 with a high emission scenario RCP8.5. The multi‐model median changes relative to the period 1986–2005 and signal‐to‐noise ratios for the annual warmest night (TNx, units: °C) over China associated with a 2 °C global warming as simulated by 29 CMIP5 models under RCP8.5. More than 80% of statistically significant models agree on the sign and signal‐to‐noise ratio is greater than 1.0, which is presented in colour with solid circles. More than 80% of statistically significant models agree on the sign and signal‐to‐noise ratio is less than 1.0, which is presented in colour with hollow circles. Less than 50% of models have a statistic