Anthropogenic Signal Detection Of Extreme Precipitation In China Based On Field Significance Test

Signal detection of regional extreme precipitation is of great significance for understanding the impact of global warming on extreme precipitation.However,there are still great uncertainties in the conclusion.This paper used a variety of observed data to estimate the response of extreme precipitation in China to global mean surface temperature using a more robust statistical method,and used the idea of field significance test to determine whether observed changes of extreme precipitation can be explained by internal variability from three aspects.The results of Coupled Model Comparison Plan phase 6(CMIP6)were used to further verify this conclusion after determining that anthropologic influence can be detected in observed extreme precipitation.Meanwhile,the time of emergence when the first human activities signal could be detected in extreme precipitation was determined,and the factors affecting the uncertainty of conclusion were analyzed.The frequency of extreme precipitation in the middle and the end of this century is also estimated.The main conclusion are as follows.With global warming,there is an overall intensification of extreme precipitation in China at a rate similar to that predicted by the Clausius-Clapeyron equation.Regional differences in the response of extreme precipitation to global average temperature were significant,showing a “wetting in the south and drying in the north”pattern.The southeast and northeast regions showed a positive correlation and significant response while some regions such as Northeast China,North China and Southwest China showed weak response or even negative correlation.The field test results of three aspects all showed that the observed extreme precipitation changes were very different from the internal variability and cannot be explained by it.CMIP6 had a good simulation of extreme precipitation in the northwest but poor simulation in the southeast and generally underestimated the observed changes.The field test based on CMIP6 data showed that simulated historical extreme precipitation cannot be explained by internal variability.This conclusion was further confirmed by the internal variability simulation of the Detection and Attribution Model Intercomparison Project(DAMIP).The observed signal of extreme precipitation changes could be detected at the beginning of this century and the signal in Rx1 ay was stronger than Rx5 day.The results of models showed that under high emission scenario,the signal of multi-model mean could be detected at the beginning of this century and would be detected in all models in 2040 s.Many factors affected the estimation of internal variability and the spatial coverage rate and spatial dependence had a great influence.In the middle and end of this century,the spatial median of Rx1 day in China increased by 14.5% and 28% on average compared with 1961-1990 periods.The longer return period,the more frequent the precipitation would be in the future.In the middle and end of this century,the probability of extreme precipitation occurring once in 50 years increased by 3.1 times and 6.3 times,respectively,to once in 16 years and once in 8 years.The counterpart of once in 100 years increased by 3.5 times and 6.8times,respectively,to once in 29 years and once in 12 years.