| With global warming,there are obvious regional variations in extreme hourly precipitation(EXHP)trends over China during the past half century.The contributions from changes of tropical cyclone(TC)and the extensive urbanization are unclear.Using multiple datasets including the gauge-based hourly precipitation dataset and the TC best track data,this paper analyzes the contribution of TC and the urban-rural differences of EXHP changes over eastern China(i.e.,east of 96°E)during the warm seasons of 1975 to 2018.The Tropical cyclones(TC)precipitation is objectively identified firstly.The total precipitation(ALL)is divided into the TC precipitation and non TC precipitation.The EXHP is defined using the 95th percentile of cumulative density function hourly precipitation during30 years(1975-2004)as the threshold value.The TCs influencing China during the analysis period are divided into three groups according to the number of EXHP records produced by each TC.If a TC produces the highest 10%EXHP records(i.e.,at least 400),it belongs to the A-group;If a TC produces EXHP records less than the average of all TCs(160),it belongs to the C-group;Otherwise,it belongs to the B-group.The results show that,in terms of the climatological characteristics,the duration of the TC influence period exhibits the largest difference among the groups,followed by the TC peak intensity,with insignificant difference in TC moving speed during its influence period.Moreover,the spatial coverage and high-frequency area of EXHP produced by the C,B,and A groups advance toward northwest,which is mainly related to the extent that they reach inland.In terms of the long-term trend,the total occurrence frequency of tropical cyclones shows a downward trend,while the total frequency of TC-EXHP shows a significant upward trend that is mainly contributed by the group A.Moreover,the EXHP generated by tropical cyclones increases significantly in the eastern part of the Yangtze River Basin during 19750-2018,and its trend is much higher than those over other regions in China.The contribution of TC to the increase of total precipitation in the Yangtze River Valley is about 35%,which is consistent with the newly found long-term trend of the shift of TC in the West Pacific toward the East China Sea and the eastern part of China.These results indicate that the change of tropical cyclones over the tropical ocean has a certain contribution to the spatial difference of the EXHP trend in China.Then,two different methods are used to construct the time series of urban and rural precipitation,respectively,to analyze the difference of EXHP changes between urban and rural areas.In one method,a station is identified as city or non-city based on the fraction of built-up area within 2 km radius centered at the station.Such urban-rural identification is made for each of seven subperiods during 1975-2018 considering the change of underlying surface properties due to urbanization.The precipitation in each 5°×5°grid is averaged at the identified urban and rural stations,respectively.A long-term series of urban and rural precipitation in each grid is thus constructed.Using this method,an urban rainy island(URI)effect is observed at less than half(43%)of the grids,where the trend of EXHP is greater at cities than that at non-city stations.In contrast,an urban heat island(UHI)effect is found at much more grids(89%)where the 2-m air temperature increasing with a larger magnitude at cities than non-cities.In the other method,the overall distributions of land use/land type after the rapid urbanization,coastline and terrain are collectively considered to identify a key region of urban agglomeration,which includes multiple cities and nearby areas under their possible influence.Taking a 4°×4°grid of the Yangtze River Delta(YRD)area as an example,using this latter method the proportion of urban stations with significant EXHP increase(81%)is higher than that of rural stations(24%),and the increase trend of the city-average EXHP is significantly higher than that of the rural(34.8 mm·decade-1vs.11.7 mm·decade-1),i.e.,a clear URI effect is found.A clear UHI effect is also found with this latter method,i.e.,0.41?·decade-1vs.0.34?·decade-1.However,using the former method,a URI effect is not found in the YRD area,as the EXHP increase trend at the rural stations slightly exceeds that at the cities,while a clear UHI effect is still seen(the 2m air temperature trends are 0.45?·decade-1and 0.29?·decade-1at the cities and rural stations,respectively).These results indicate that to explore the impact of urbanization on extreme precipitation,it is necessary to collectively consider specific geographical environment and climatic background(such as low-level prevailing wind direction)in the identification of city and rural stations,especially for a large-size urban agglomeration as those in eastern China. |
PDF Full Text Request