EGI OpenIR
中国天山雪深重建与陆面积雪过程模拟
Alternative TitleHistorical snow depth reconstruction and improvement in snow process in land surface model over the Tianshan Mountains,China
李倩
Subtype博士
Thesis Advisor李兰海
2019-06-30
Degree Grantor中国科学院大学
Place of Conferral北京
Degree Discipline理学博士
Keyword积雪资源 雪深重建 WRF/Noah-MP 模式 Snow resource Snow depth reconstruction WRF/Noah-MP model
Abstract水资源是影响干旱区生态环境和制约社会经济可持续发展的关键要素,中亚干旱区地表水资源匮乏,但季节性积雪资源丰富,作为中亚水塔的天山具有丰富地积雪资源。 中国天山山区气象站点稀少、分布不均匀, 遥感积雪产品时间序列短且精度有限,陆面模式在融雪期的模拟存在较大偏差,这些不足使得针对区域积雪长期变化的评价和高精度积雪数据的获取比较困难。因此揭示中国天山雪深在过去百年具有怎样的变化规律,以及提高陆面模式对于中国天山山区雪深和雪水当量的模拟精度尤为重要。 本论文研究利用气象站点数据,分析中国天山地区积雪各参数的时空变化规律及其分别在积雪累积期和消融期的影响因素。结合再分析数据产品,应用人工神经网络(Artificial Neutral Network, ANN)等方法,重建中国天山地区百年雪深序列。利用遥感反照率产品,修改区域气候模式(Weather Research and Forecasting Model, WRF) /Noah-MP 模式中的陆面积雪过程,提高融雪期雪深和雪水当量的模拟精度。有效地利用现有数据,促进积雪过程的研究,改进与发展陆面模式的积雪过程,获得精度更高、范围更广泛的积雪数据,是干旱区水资源评估的科学基础,也是区域积雪变化研究的迫切需要。本文的主要结论如下:(1) 中国天山积雪在 1961–2015 年的空间分布和变化特征具有明显地区域差异。 积雪持续日数和最大雪深呈现出西北高-东南低的空间格局,但是积雪消融率的空间分布则与前两者相反。在 1961-2015 年间, 中国天山地区的积雪平均于 11 月 24 日出现、 1 月 27 日达到最大、 3 月 18 日消失,多年积雪持续日数为115 天。 在 1961–2015 年间, 积雪持续日数以 4.4 天/10a 的速度减少, 积雪持续日数的变化,有 63%是由积雪结束日期的提前所导致, 37%是由积雪出现日期的推后所导致。 积雪累积期的平均气温 0.09 °C/10a 和积雪消融期的有效累计正积温 1.15 °C/10a 在 1961-2015 年间均出现了增大趋势,从而分别导致了积雪出现日期的推后和积雪结束日期的提前。 中国天山地区的雪深在这 54 年间出现了显著增加的趋势,增加率为 2.08 cm/10a。(2) 在 1901–1960 年间, 通过分析校正后年平均气温和降水数据, 发现中国天山气温和降水均出现了增加趋势,但是仅年降水的变化趋势通过了显著性0.05 水平检验。 所重建的历史雪深(1901–1960 年) 的空间格局, 与站点观测的空间分布特征一致,呈现西北-东南递减的分布特征。 中国天山地区总体雪深经历了增加的趋势,但是该趋势并不显著。在垂直方向上,雪深的变化趋势也体现出了一定的空间特征。重建历史雪深的增加趋势随着海拔的升高,出现了增大的趋势, 其中 1000 m–1500 m 海拔带变化率最小。 中国天山地区南坡和东部的雪深在历史时期 1901–1960 年、 观测时期 1961–2014 年、 总体时期 1901–2014 年均出现了增长趋势。然而, 北坡、伊犁河谷、天山整体的雪深在总体时期出现了下降趋势,而它们在历史时期和观测时期分别出现了上升趋势。雪深在三个研究时段中呈现出了不同的趋势,说明雪深的变化呈现出了年代际的特征,也体现了历史雪深重建的重要性和必要性。(3) 通过对 2014 年 10 月至 2015 年 4 月整个积雪季的模拟, 比较 Noah、Noah-MP 两套方案与站点气温、雪深、雪水当量数据, 结果表明 Noah-MP1 方案的模拟精度最高, Noah-MP2 方案次之, Noah 方案最差。与站点观测数据相比,整个积雪季三套方案的模拟精度结果显示气温全部为低估,雪深和雪水当量均为高估。 利用 MOD10A1 积雪反照率产品和 MCD43A4 地表反照率产品,在积雪消融期对 Noah-MP 模式中的反照率进行替换。通过分析结果发现两种改进方案均能提高模式的模拟能力。 相比 Noah-MP 两套改进前的方案,两套改进后方案的气温、雪深、雪水当量结果均更加接近站点观测数据。 其中,利用积雪反照率 MOD10A1 积雪反照率产品改进模式方案的结果精度更高。
Other AbstractWater resource is the key factor to ecological environment and community inarid areas. It is abundant in seasonal snow resource in Central Asia, where is lack ofsurface water. As the water tower in Central Asia, the Tianshan Mountains areabundant in snow resource. The meteorological stations are scarce and theirdistributions are heterogeneous in the Tianshan Mountains. The remote sensingretrievals are in short duration record, and their resolution and accuracy are limited.The snow data simulated by land surface models are in low accuracy. These issueslimit the assessment of long-term variation in snow and the demand about snowdataset with high accuracy. Therefore, it is important to reveal the historical variationin snow during last century and to improve the simulating ability in snow depth andsnow water equivalent of land model surface in the Tianshan Mountains, China. Thisthesis applied daily snow depth, temperature, and precipitation during 1961 to 2015from in-situ observations to analyze the distribution and variation in snow metrics thatrelated to snow depth, as well as their attributions during snow accumulation andablation period, respectively. The historical snow depth series from 1901 to 1960 werereconstructed by Artificial Neutral Network (ANN) method based on in-situobservations and reanalysis dataset. To improve the simulating accuracy in snowdepth and snow water equivalent during snow ablation season, this thesis modified thealbedo in the Weather Research and Forecasting Model (WRF) /Noah-MP by remotesensing albedo products. The results showed:(1) It has a distinct snowpack patterns within the Tianshan Mountains, China.The results showed that the snow depth maximum and snow cover duration exhibitedsimilar distribution, i.e., decreasing in pattern of a northwest-to-southeast direction.However, the distribution of snow ablation slope was by contrary to that of snowdepth maximum and snow cover duration. The snow occurred on November 24th,maximized on January 27th, and disappeared on March 18th during the period from1961 to 2015. The mean snow cover duration was 115 days during 1961–2015.Significant trend toward a shortened snow cover duration was identified by 4.4 daysper decade. The contribution analysis revealed that the changes in snow coverduration was primary the result of earlier snowmelt which contributed at 63%, ratherthan later snow onset which contributed at 37%. The later snow onset day and earliersnow end day were coherent with the increase in both the temperature (0.09 °C perdecade, p<0.05) and effective >0 °C accumulative temperature (1.15 °C per decade,p<0.05) during snow accumulation season and snowmelt season, respectively. Inaddition, a significantly increased snow depth maximum was found by 2.08 cm perdecade, against the warming background in the northern hemisphere.(2) In the historical period (1901–1960), trend analysis showed that the annualtemperature and annual precipitation of the entire Tianshan Mountains, Chinaincreased during 1901–1960, only significant at 0.05 confidence level in precipitation.The results indicated that the reconstructed snow depths captured the long-termvariation and spatial distribution in the study area. For spatial scale, higher values ofsnow depth occurred in the western and northern Tianshan Mountains, while lowerones appeared in the southern and eastern Tianshan Mountains. For temporal scale,increases in snow depth were detected in the southern and eastern TianshanMountains, China during all three periods. The trends in snow depth indicated anincrease in the western, northern and overall Tianshan Mountains, China during theperiods of 1901–1960 and 1961–2014, but showed a decrease during the overallperiod of 1901–2014. The difference in variation of snow depth trends in differenttemporal scales indicates that the time scale of snow depth increase is decadal ratherthan centennial. It is important to construct historical snow depth to due to thesedifferent trends at different spatial and temporal scales.(3) Simulation was conducted based on different albedo schemes of the snowseason October 2014 to April 2015. The 2-m temperature, snow depth, and snowwater equivalent from Noah and different Noah-MP albedo schemes were evaluatedby comparing with the in-situ observations. The 2-temperature exhibitedoverestimation while the snow depth and snow water equivalent revealedunderestimation in the study area during the whole snow season. The Noah scheme showed the lowest accuracy, the Noah-MP1 scheme got the highest accuracy duringthe whole snow season. The Noah-MP scheme was modified by two remote sensingalbedo products during snow melt season, i.e. MOD10A1 and MCD43A4. The resultsshowed that the modification of albedo in land surface model could improve thereproducing procedure of the Noah-MP model. The 2-m temperature, snow depth, andsnow water equivalent showed less bias of the two modified schemes during snowmelt season compared with in-situ observations. The scheme that modified byMOD10A1 product exhibited the highest accuracy in this study.
Subject Area自然地理学
Language中文
Document Type学位论文
Identifierhttp://ir.xjlas.org/handle/365004/15295
Collection中国科学院新疆生态与地理研究所
研究系统
Affiliation中国科学院新疆生态与地理研究所
First Author Affilication中国科学院新疆生态与地理研究所
Recommended Citation
GB/T 7714
李倩. 中国天山雪深重建与陆面积雪过程模拟[D]. 北京. 中国科学院大学,2019.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[李倩]'s Articles
Baidu academic
Similar articles in Baidu academic
[李倩]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[李倩]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.