On the Emergence of Anthropogenic Signal in Extreme Precipitation Change over China

The detection of anthropogenic influences on climate extremes at regional scale is not only important for the development of national climate change policy, but also the hot topic and grand challenge in regional climate change research. We use non-stationary GEV based on r-largest events to estimate the extreme precipitation change due to anthropogenic influence in each station or grid cell. Field significance resampling-based test was used to determine whether the anthropogenic influence is significant in the overall region. Global climate simulations from CMIP5 under the RCP8.5 scenario are used to examine the time at which an anthropogenic influence becomes detectable in extreme precipitation over China and the change in probability of extreme precipitation with certain magnitudes when the changes are detectable. Results shown that anthropogenic influence is not significantly detected over China in the observational record or simulations from 1961 to 2012 based on the test of field significance. Simulations indicate that such change would become detectable in the future by around 2035 under RCP8.5 scenario. Large changes would already manifest by the time of signal detection; for example, extreme precipitation events that occur on average once every 20, 50 and 100 years in current (1986–2005) climate would reduce to about 15, 34, and 63 years on average by the time of detection around 2035.