| Recent years,a few of international research institutions have developed their respective independent datasets with gradual improvements and updates.These different datasets,which are used to establish the anomaly series of global land surface air temperature,are generally consistent in producing the long-term temperature variation.And these results for variation characteristics have been unanimous recognized.However,there are also some problems in these datasets,such as low coverage,short record length and so on.To solve these problems,a new CMA-LSAT vl.0 dataset was developed at the National Meteorological Information Center(NMIC)of China Meteorological Administration(CMA),by integrating the multiple datasets including those of Asian countries,South America and Africa region through the exchange of data.In this study,the CMA-LSAT vl.0 dataset was adopted to analyze the global and continental land surface air temperature trends for varied periods of 1901-2015,and the uncertainty of the global surface temperature also is analyzed.These conclusions could deepen the understanding of the hot issues in global warming research.The main conclusions are as follows:(1)Two obvious temperature-rising periods occurred from 1901 to the late 1930s and from the early 1980s to the mid-2000s.Meanwhile,two warming slowing periods or cooling periods occurred,with the first between the early 1940s and the mid 1970s,and the second during the last 18 years.During the period 1901-2015,the global land surface annual mean air temperature increased at a rate of 0.104?/decade,and the Northern and Southern hemispheres witnessed annual mean warming at 0.115?/decade and 0.088?/decade respectively.The annual warming in the Arctic,as well as the high latitudes of the northern hemisphere,was larger and more significant than that in the mid-low latitudes.The zonal mean temperature trends showed that the slowdown of climate warming during the recent 18 years was mainly the results of the stagnation or even reversal of the land surface air temperature increase in the mid-low latitudes.The seasonal mean warming trends in the northern hemisphere were significantly higher than those of the southern hemisphere.The winter and spring seasonal mean warming trends in the northern hemisphere were higher than those of the summer and autumn seasons.In 1901-2015,Asia and Europe registered the largest warming trends of above 0.110?/decade,while Africa and Oceania experience the smallest warming of only 0.074-0.084?/decade.During the periods 1951-2015 and 1979-2015,the continents all showed more intensive warming trends,which were particularly obvious in Europe and Asia.In the last 18 years,with the exception of the Antarctica and Oceania,the continents showed different degrees of slowdown of climate warming or even a cooling trend.(2)In the period 1901-2014,results indicate that the global land surface DTR significantly decreases at a rate of-0.036?/decade over the 1901-2014 period,mainly due to the large decrease in DTR from 1951 to 2014.For the first half of 20th century,most grid boxes show a positive DTR trend,with the positive trends of 32.4%grid boxes being statistically significant,leading to a large and significant increase of 0.048?/decade in DTR.However,a dramatic reversal in DTR change occurs around 1951,with most parts of global lands seeing a shift from increasing to decreasing trends.Global land average DTR decrease during 1951-2014 is-0.054?/decade,with 45.0%grid boxes showing negative trends with statistical significance.The reversal phenomenon is more obvious in the Northern Hemisphere than that in the Southern Hemisphere.For the periods 1979-2014 and 1998-2014,the decreasing trends in DTR merely occur in the Northern Hemisphere.The Southern Hemisphere experiences larger increases during the two recent periods.Asia,eastern North America and Australia exhibits widespread decreasing in DTR although the trend pattern for global DTR is generally mixed during 1979-2014 and 1998-2014.There is a good negative correlation between DTR and precipitation in the Northern Hemisphere from 1901 to 2014,with a correlation coefficient of-0.61.The change in precipitation and number of volcanic eruptions,and the "early brightening" of Europe(Stockholm station)all benefit the increase of DTR at global and regional scales in the first half of the 20th century.(3)In the period 1951-2014,the warming trends in northeast China for Tmax,Tmin and Tmean are 0.20?/decade,0.42?/decade and 0.34?/decade respectively for the whole area,with the warming rate of Tmin about 2 times of Tmax,and the upward trend of Tmean obviously higher than mainland China and global averages;In the period 1998-2014,the annual mean temperature consistently exhibits a cooling phenomenon in Northeast China,and the trends of Tmax,Tmin and Tmean are-0.36?/decade,-0.14?/decade and-0.28?/decade respectively;seasonal mean cooling mainly occurs in northern winter and spring,but northern summer and autumn still experience a warming,implying that the annual mean temperature decrease is controlled by the remarkable cooling of winter and spring;Compared to the global and mainland China averages,the hiatus phenomenon is more evident in Northeast China,and the cooling trends are more obvious in the cold season;The Northeast China cooling trend occurs under the circulation background of the negative phase Arctic oscillation,and it is also closely related to strengthening of the Siberia High and the East Asian Trough,and the resulting the stronger East Asian winter monsoon over the GWH period.(4)From 1901-2015,the Hindu Kush Himalayan region,all of Tmax,Tmin and Tmean,show sensitive to elevation.Tmean increasing rate is 0.029?/decade per 500m,the increasing rate of Tmean about 2 times of Tmax,and the trend of Tmin increased by 0.023?/decade per 500m.The three elements in the whole season showed sensitive to elevation,the EDW phenomenon is more obvious in winter.In the HKH region,the change in extreme cold events in the HKH appears to be more sensitive to elevation,whereas the change in warm extremes shows no detectable relationship with elevation.Frost days and minimum Tmin also have a good relationship with elevation,and the trend in frost days decreases with an increase in elevation while the trend in minimum Tmin increases with an increase in elevation.Compared with the previous research results,the overall results are consistent basically,different is that that previous study neglected the Tmax and Tmin in spring,autumn and winter also exist EDW phenomenon,but also between the trend of cold event index and DTR also sensitive to elevation.(5)It is found that,the urban bias in the annual mean land surface air temperature anomaly series of Northern hemisphere is 15.6%between 1951-2015,the contribution rate in Europe and North America is similar with that of the Northern Hemisphere average values.In Asian large-scale fast city area,urban bias contribution rate is 19.3%.It is also found that the urban bias in the annual mean land surface air temperature anomaly series of mainland China has led to an overestimate of at least 2.8%for the global land surface annual mean air temperature trends for the period 1951-2014.The level of uncertainty caused by the poorly early data is about 3%-15%.In addition,the level of uncertainty caused by the difference regional average method and the trend estimation method of different regions is 4%-9%and 1%-6%,respectively.Therefore,the influence of different statistical methods on large scale long term trend estimate is lower than the trend magnitude. |
PDF Full Text Request