Observed Trend Changes In Extreme Temperature Over The Global Land,1951-2018

The global surface air average temperature is increasing rapidly over the last 100 years.In this context,the global land extreme air temperature changes in the frequency,intensity and duration,which direct important impact on human and natural ecosystems.A better understanding of long-term change of extreme climate thus cannot only contribute to extreme temperature change detection,attribution and projection,but can also provide the scientific and technological support for global and regional climate change adaptation and the meteorological disaster risk management.At present,there have been some studies of global land extreme temperature in the world,but most of these studies are not based on the homogeneous adjusted daily air temperature dataset.Recently,some national and regional homogeneous adjusted daily air temperature datasets have been developed.Based on these latest homogeneous datasets,a new global land daily surface air temperature dataset was integrated,and then the spatiotemporal characteristics of global land extreme air temperature over 1951-2018 were studied using the indices recommend by the Expert Team on Climate Change Detection and Indices;the global land rural stations are selected using the machine learning method based on the United States Climate Reference Network(USCRN)and global land use/land cover(LULC),and then the urbanization contribution to the observed trend in extreme temperature indices series are evaluated.The main research results are summarized as follows:1)A new global land daily air temperature dataset was integrated.The new dataset is based on the Global Historical Climatology Network-Daily Database(GHCND)as a benchmark,and integrates the European Climate Assessment & Dataset(ECA&D),Australia,Canada and China's homogeneous daily surface air temperature data set.After the dataset was integrated,a consistent quality control was performed again,and then the inhomogeneous test was performed by using the RHtests software,only the temperature series without step changes at the high confidence level are retained.The newly integrated dataset contains the most known homogenous adjusted daily surface air temperature dataset,which reduces the bias of the research results caused by the inhomogeneities of the data;meanwhile,the newly integrated dataset has an improved in spatial coverage in the early(1951)and late(2018).2)The study revealed the spatiotemporal characteristics of the global land extreme temperature change trend from 1951 to 2018.The main conclusions are concluded that:(i)the global land annual and seasonal average extreme temperature series all experienced significant warming trend over the study period 1951-2018,i.e.cold threshold indices(frost days,icing days,cold days and cold days)decreasing and warm threshold indices(summer days,tropical nights,warm days and warm nights)increasing.The extreme temperature indices based on the daily minimum temperature generally had a stronger and more significant trend than those based on daily maximum temperature.(ii)The most significant warming in most extreme temperature indices occurred after the mid-1970 s,and before the mid-1970 s the global land average indices series generally showed no significant change.(iii)In the central and eastern regions of the United States,the extreme temperature indices derived on the daily maximum temperature generally did not experienced a warming trend,forming a so-called "warm hole" phenomenon.(iv)The volcanic eruption has a significant impact on the warm extreme indices,especially the warm extreme indices derived from the daily maximum temperature.In the year of the volcanic eruption and the following 1-2 years,the frequency of warm extreme events decreased significantly.3)The urbanization effects in the extreme air temperature series are quantitatively evaluated.The percentage of urban areas within 1-12 km buffer radius around the USCRN stations are used as the training dataset,the method of machine learning was used to select rural stations from global stations that are similar to the surrounding environment of USCRN stations,and then the global/regional average time series for rural reference stations and all stations are calculated,respectively;finally,the urbanization effects of global land and regional average extreme temperature indices time series are quantitatively evaluated.The main conclusions are summarized as follows:(i)the significantly urbanization effects are detected in most of the extreme temperature indices series over the global land during 1951-2018,and the most significantly urbanization effect are detected in the extreme temperature indices that derived from the daily minimum temperature.(ii)The detected urbanization effect mainly occurred after the mid-1980 s,which is possible related to the accelerated development of the world economy after the 1980 s.(iii)There are significant differences in the urbanization effects of different regions,with East Asia experienced the most obvious urbanization effect,and Europe and North America experienced the weakest urbanization effect.