【作者】 刘芳；

【导师】 徐长春；

【作者基本信息】 新疆大学 ， 地理学， 2022， 硕士

【摘要】 中国西北干旱地区是对全球气候变化最为敏感的地区之一,正确认识该地区未来气候变化对于区域经济社会发展和防灾减灾具有重要意义。长期以来,耦合模式相互比较项目一直是气候变化模拟与预估的重要工具。随CMIP6的开展,新的全球气候模式数据已经完全公布。评估其在西北干旱区的模拟能力并预估其未来变化可以为政府及决策者提供相应的参考,提前规避因气候变化带来的风险。本文基于CMIP6中31个模式和CN05.1格点资料的平均气温、最高气温、最低气温和降水数据,评估了CMIP6模式和四种模式集合(MME、MMM、PME和PMM)对中国西北干旱区1961-2014年气候的模拟能力,并基于未来情景预估数据对该地未来可能的气候变化进行预测。主要结论如下:(1)1961-2014年,西北干旱区三种气温要素均呈显著的增温趋势,且增温速率表现为最低气温(0.43℃/10a)>平均气温(0.30℃/10a)>最高气温(0.22℃/10a),空间上表现为北疆、南疆南部和河西走廊东部升温速率最快。大部分模式都可模拟出气温的变化趋势且与观测数据的相关性较好,但平均气温和最高气温模拟的温升速率大于观测速率,而最低气温较观测则整体偏低。四种模式集合在相关性、趋势变化和气候平均态的模拟上均优于单个模式,且PME和PMM优于MME和MMM,其中PME表现最好。(2)近54年来,西北干旱区降水以4.99 mm/10a的速率显著增加,大多数模式(29个)能够模拟出降水的增加趋势,但模拟强度偏低;空间上,在新疆中天山、南疆西部和祁连山附近降水增加最显著。相关性和空间气候平均场上、模式的模拟效果均较气温差。MME、MMM、PME和PMM在相关关系、趋势和气候平均态上表现均比单个模式好,且PME和PMM可有效的减小趋势和多年降水在山区附近的偏差分布范围,但综合来看PME的模拟能力最好。(3)历史时期,对于气温,经一元线性回归和气候漂移订正后,PME能准确的模拟出平均气温、最高气温和最低气温的年内和年际变化特征,空间上,两种方法将模式与观测之间的偏差缩小了10倍;对于降水,经一元对数回归订正与扣除气候漂移后,PME与观测的时空相关系数分别为0.79和0.98(P<0.01),两种方法分别将空间偏差缩小了约24.5%和93.3%。总体来说,扣除气候漂移对三种气温要素和降水的订正效果较好。(4)2015-2100年西北干旱区平均气温、最高气温、最低气温在4种情景下时间和空间上都呈显著的增温趋势,增温速率表现为SSP1-2.6<SSP2-4.5<SSP3-7.0<SSP5-8.5;三种气温要素中,最低气温升温速率最快,但其变化趋势的空间分布格局基本一致,升温速率都表现为从北向南递减。相对于基准期,三种气温要素在21世纪近期、中期和末期的四种排放情景下升温幅度不同,近期四种情景下增温幅度之间差异不大,而中期和末期随辐射强迫增强,增幅变大,空间上,平均气温、最高气温和最低气温的增温幅度均表现为从北向南逐渐减弱,且升温大值区均在北疆北部。(5)2015-2100年西北干旱区降水在四种情景下总体呈增加趋势,但SSP1-2.6情景下降水变化不明显;空间上,SSP1-2.6情景下在南疆中部、新疆北部和东部以及河西走廊北部地区降水呈下降趋势,但未通过显著性检验,而SSP2-4.5、SSP3-7.0和SSP5-8.5情景下,整个西北干旱区的降水呈增加趋势,且山区的降水增速大于盆地,随着时间的推移和排放情景的增加,降水增速也越大。相对基准期,降水在21世纪近期、中期和末期的增幅范围为4.56～42.11%,辐射强迫越强,增幅越大;空间上,除昆仑山区附近以及且末县东南角降水较基准期减少外,其余地区降水总体增加,增幅较大的地区为南疆南部,河西阿拉善地区以及阿克苏地区。更多还原

【Abstract】 The arid region of northwest China is one of the most sensitive areas to global climate change.A proper understanding of future climate change in this region is of great significance for regional economy and social development as well as disaster prevention and alleviation.The Coupled Model Intercomparison Project(CMIP)has long been an important tool for climate change simulation and prediction.With the development of CMIP6,the new global climate model data have been fully published.Evaluating its simulation capability in the arid region of northwest China and predicting its future changes can provide governments and decision makers with corresponding references to avoid the risks caused by climate change in advance.Based on the mean temperature,maximum temperature,minimum temperature and precipitation data from CMIP6 and CN05.1,this paper evaluates the simulation ability of individual models and four model-ensembles(MME,MMM,PME and PMM)for climate change in the arid region of northwest China during 1961-2014,and predicts possible future climate change under different scenarios.The main findings are as follows.(1)From 1961 to 2014,the average,maximum,and minimum temperatures in the arid region of northwest China showed significant warming trends,and the warming rates behaves as minimum > average > maximum.Spatially,northern Xinjiang,southern Xinjiang and the eastern part of Hexi Corridor have the fastest warming rate.Most of the models can simulate temperature trends well and have good correlation with the observed data,but the simulated average and maximum temperatures have a greater warming rate than the observed values,while the minimum temperatures are overall lower than the observation.The four model ensembles outperform the individual models in terms of correlation,trend change,and climate mean state,and PME and PMM outperform MME and MMM,with PME performing best.(2)In the past 54 years,precipitation in the arid region of northwest China has increased significantly at a rate of 4.99 mm/10 a.Most models(29)were able to simulate the increasing trend of precipitation,but the simulation intensity was low;Spatially,the precipitation increases most significantly in the middle Tianshan Mountains,the west of southern Xinjiang and t near the Qilian Mountains.In terms of correlations and spatial climate mean fields,the simulation effect of the model on precipitation is worse than that of air temperature.MME,MMM,PME and PMM are better than single model in correlation,trend and climate average state,and PME and PMM can effectively reduce the deviation distribution range of trend and precipitation in mountainous areas for many years,but PME has the best simulation ability.(3)In the historical period,for air temperature,PME can precisely simulate the intra and inter-annual change of mean,maximum and minimum temperatures after the revision by one-dimensional linear regression and climate drift.And spatially,the deviation between the model and the observation is reduced by 10 times by the two methods;For precipitation,the spatial and temporal correlation coefficients between PME and observations were 0.79 and 0.98(P<0.01),which reduced the spatial bias by about 24.5% and 93.3% by the two methods,respectively.Generally speaking,the correction effect of climate drift on three temperature elements and precipitation is better.(4)In 2015-2100,the average,maximum,and minimum temperatures in arid region of northwest China show significant warming trends in time and space under the four scenarios,and the warming rates show SSP1-2.6 < SSP2-4.5 < SSP3-7.0 < SSP5-8.5.Among the three temperature elements,the warming rate of the minimum temperature is the fastest,but the distribution pattern of the spatial trends of the three temperature elements is basically the same,and the warming rates show a decreasing distribution pattern from north to south.Relative to the base period,the three temperature elements differ in their warming magnitudes in the four emission scenarios for the near,mid and late 21 st century,with little difference between the warming magnitudes under the four near scenarios in the near period,while the increase becomes larger with enhanced radiative forcing in the mid and late periods.Spatially,the warming of the mean,maximum and minimum temperatures show a gradual weakening from north to south,and the areas of large values of warming are all in the northern part of northern Xinjiang.(5)During 2015-2100,the precipitation in the arid region of northwest China showed an increasing trend under four scenarios overall,but the change of precipitation under the SSP1-2.6 scenario was not significant.Spatially,precipitation in central southern Xinjiang,northern and eastern Xinjiang and the northern Hexi Corridor shows a decreasing trend under the SSP1-2.6 scenario,but it does not pass the significance test,while the SSP2-4.5,SSP3-7.0,and SSP5-8.5 scenarios show an increasing trend in precipitation in the entire Northwest Arid Zone,and the rate of precipitation increase in mountainous areas is greater than that in basins.With the passage of time and the increase of emission scenarios,the increasing rate of precipitation will increase.Relative to the base period,precipitation increases in the range of 4.56 ～ 42.11% in the near,mid and late 21 st century,and the stronger the radiative forcing,the greater the increase.Spatially,except for the precipitation near the Kunlun Mountains and the southeast corner of the Heimei,which decreased compared with the reference period,the precipitation in the rest of the regions increased in general,and the regions with larger increases were the southern part of South Xinjiang,the Alashan region in the west of the river and the Aksu region.更多还原