EGI OpenIR  > 研究系统  > 荒漠环境研究室
中亚积雪时空分布的变化及其对关键气候要素的响应
刘金平
学位类型硕士
导师包安明
2015
学位授予单位中国科学院大学
学位授予地点北京
学位专业地图学与地理信息系统
关键词积雪 再分析数据 时空变化 中亚地区 影响因子
摘要积雪是冰冻圈要素之一,对气候变化响应敏感,同时影响水文循环系统。根据IPCC第五次评估报告第一工作组报告指出全球气候系统变暖是明确无误的,再一次证实全球正在变暖的事实。在这个大背景下,本文根据ERA-Interim数据提取出雪盖信息,利用实测雪深数据和MODIS积雪产品验证了精度,在此基础上,计算出1987~2014年各年的积雪初日和积雪终日并分析了分布特征,以年内和年际变化着手分析了中亚地区积雪在时间上的变化特征,利用EOF方法和R/S分析法分析了积雪在空间分布上的变化特征及预测了中亚地区积雪未来变化的趋势,研究了中亚地区的气温、降水、海平面压力、风速等近30年的变化,利用偏相关分析研究了影响中亚地区积雪的关键气候因子。主要研究内容及结论如下: (1)通过ERA-Interim再分析数据提取了中亚地区的雪盖信息,并利用实测雪深数据和MODIS积雪产品验证,结果显示相比实测雪深,ERA-Interim雪深精度为72.1%,对比MODIS积雪产品,积雪覆盖率的平均绝对误差为3.32%,均方根误差为0.042,满足本文研究积雪变化的需要。 (2)中亚地区积雪在10月份开始积累,积雪最大值主要出现在1月和2月,最小值集中在10月和4月,相比于积雪面积最大值,最小值波动较小,以积雪面积最大值和积雪面积最小值都处于减少的趋势但不显著。 (3)高纬度地区、山区的周边地区在近30年间积雪初日和积雪终日分别具有延后和提前一个月的趋势,低纬度和低海拔地区的积雪期一直较短,在近几年,处于低纬度地区的乌兹别克斯坦、土库曼斯坦部分地区的积雪逐渐消失并具有从中亚西南部向东北延伸的趋势。总的来讲,研究区内积雪初日延后而积雪终日提前,导致积雪期缩短。 (4)积雪在空间分布上具有极大的差异性,山区和高原处于最高,其次是哈萨克斯坦北部,除山区和高原外,整个中亚地区的积雪频率是纬度的函数,随着纬度的升高频率增加;以大寒日进行分析积雪的空间变化特征,结果表明中亚地区的积雪面积在0.05显著性水平上减少,这种趋势是从2005年开始并在2009年显著。 (5)中亚地区积雪变化的关键异常区主要分布在昆仑山区、天山南坡,其次为哈萨克斯坦的西南部地区,利用EOF分析方法并结合Mann-Kendall突变检验得到中亚地区的积雪在近30年发生了突变,从2000年开始并在2005年显著,最明显变化的区域为山区和哈萨克斯坦的积雪由异常多变为异常少,EOF第二模态显示低海拔和高海拔地区呈反相变化;通过R/S分析对中亚地区的积雪进行了预测。 (6)通过偏相关分析,气温、降水、风速、海平面压力等气候因子对中亚积雪影响的大小为:气温>风速>降水>海平面压力,从空间分布来看,高纬度地区的关键影响因子是风速,气温的影响范围最大,占据了整个中亚的62.9%,降水和海平面压力影响面积较小。总体而言,气温、风速、降水的分别是低海拔地区、高纬度地区、高海拔地区的主要驱动力。
其他摘要Snow is one of essential factors in cryosphereand snow is very sensitive to climat change and meantime snow can affect hydrological cycle system According to the fifth assessment report of IPCC, global warming of the climate system is unequivocal, which reconfirmed that the fact of the world is warming. Under the background, the present study extrated snow-cover data using ERA-Interim and verified the results’ accuracy by MODIS snow products. Then, we calculated the annual snow initial date and ending date from 1987 to 2014 and obtained snow distribution patterns. The characteristics of snow temperal changes in Central Asia were inveatigated by calculating the changes within year and changes in inter-annual. The characteristics of snow spatial changes were analyzed using EOF and R/S methods and then the development trends of snow changes were estimated. Finally, the changes in past 30 years of the North Pacific Oscillation and Central Asian climates variables, such as air-temperature, precipitation, sea level atmospheric pressure, wind speed and so on, were investigated. The key climate factors that affected the snow changes in Central Asia were investigated using partial correlation method. The main contents and conclusions are as follows: (1) Snow cover data was calculated from ERA-Interim data, and the results were validated using measured snow depth data and MODIS snow product. The results showed that the accuracy of snow depth data extracted from ERA-Interim is about 72.1% compared to measured data, and the average absolute error of snow cover is 3.32% with RMSE of 0.042 compared to the results of MODIS snow product. The results can satisfy snow change analysis. (2) Accumulation of snow in Central Asia started in early October, reaching to the maximum in January and February, and the minimum in October and April. The fluctuations of minimum snow cover data are less than these of the maximum. Both maximum and minimum had a tendency of decreasing according to regression analysis, but not significant. (3) In the past 30 years, the initial dates and the ending dates of snowing in high latitude areas and surround mountain areas have the tendency of delaying one month and advancing one month, respectively. Snow cover durations are always short in low latitude areas, and snow cover began to disappear in the some regions of Uzbekistan and Turkmenistan, extending from southeast of Central Asia to northeast. (4) Snow cover had great differences in spatial distributions, with the maximum appeared in mountains and plateaus and followed by the north of Kazakhstan. Except for mountain areasand plateaus snowing frequency is strongly correlated with latitude, Studying the spatial variation of snow distribution on the day of severe cold, the results showed that the area of snow cover in Central Asia is decreasing with the significance at the level of 0.05. This phenomenon started in 2005 and became remarkable in 2009. (5) The key anomaly changes of snow cover was mainly distributed in the Kunlun mountains, the southern slope of Tianshan mountain, and followed by southwest region of Kazakhstan. Using Empirical Orthogonal Function (EOF) method and Mann-Kendall methods to validate snow cover changes in Central Asia in the past 30 years, the results showed that snow cover had mutated from 2000 to 2005. The most significant changes were happened in mountain areas, plateau and Kazakhstan. The second mode of EOF showed that the change of low and high altitude was inverted. The future variation of snow cover duration in Central Asia was predicted by rescaled range analysis. (6) Through the partial correlation analysis, the order of influence power of climate factors that impacting snow cover was temperature, wind speed, precipitation and sea level pressure in descending order. In terms of spatial distribution of snow, the key factor in the high latitude was wind speed. The influence of temperature was greatest and occupied 62.9% areas in the entire Central Asia. Overall, the main driving forces in low altitude areas, high latitude areas and high altitude areas were temperature, wind speed and precipitation respectively.
学科领域地图学与地理信息系统
语种中文
文献类型学位论文
条目标识符http://ir.xjlas.org/handle/365004/14652
专题研究系统_荒漠环境研究室
作者单位中科院新疆生态与地理研究所
推荐引用方式
GB/T 7714
刘金平. 中亚积雪时空分布的变化及其对关键气候要素的响应[D]. 北京. 中国科学院大学,2015.
条目包含的文件
条目无相关文件。
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[刘金平]的文章
百度学术
百度学术中相似的文章
[刘金平]的文章
必应学术
必应学术中相似的文章
[刘金平]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。