Alternative TitleStudy of drought characteristics in the arid zones of Central Asia over the last 50 years
Thesis Advisor包安明 ; Philippe De Maeyer
Degree Grantor中国科学院大学
Place of Conferral北京
Degree Discipline理学博士
Keyword干旱 干旱特征 游程理论 三维聚类 遥感干旱监测 Drought Drought Characteristics Run Theory 3-D Clustering algorithm Remote Sensing Drought Monitoring
Abstract亚洲中部地区是具有半干旱干旱气候且水资源严重短缺的典型区域。在全球变暖和人类活动日益加剧的背景下,亚洲中部地区干旱频发,对该地区农业、畜牧业、水资源管理、经济发展及人类健康造成不可忽视的影响。干旱事件识别及干旱特征的定量化分析对于研究不同时空尺度的干旱空间格局及时间变化、制定适当的干旱灾害预防和减灾措施都十分重要。同时,干旱事件的识别及其特征的描述也是了解干旱事件的关键方法之一。然而,因为干旱和干旱特征自身的复杂性及干旱影响的广泛性以及该地区干旱监测数据的有限性导致亚洲中部地区的干旱监测研究十分有限,干旱特征尚不清楚。基于以上背景,本文旨在通过以下两个方面研究亚洲中部干旱区的干旱特征:一方面是基于标准化降水蒸散指数(SPEI) 分别使用传统游程理论方法和新提出的三维聚类算法识别研究区干旱事件并定量化描述其特征。另一方面,采用约束最优加权法和基于多遥感变量的最小误差迭代法等手段建立基于多遥感变量的农业干旱监测指标—最优比例干旱状态指数(OSDCI), 以期为亚洲中部地区的农业干旱监测提供方法支撑。本论文的主要内容总结如下:首先,在 3 个月、 6 个月和 12 个月的时间尺度上,利用游程理论和 SPEI 干旱指数识别干旱事件并分析干旱发生频次、持续时间、严重程度、强度和频发季节等。结合主成分分析(PCA)和 Varimax旋转法将亚洲中部地区划分为具有相似干旱特征的子区域;利用 Sen’s 坡度法和改良的 Mann-Kendall 方法(MMK)研究干旱趋势; 应用小波分析研究干旱周期以及干旱变化与大尺度气候模式之间可能存在的联系。结果表明: 亚洲中部地区的干旱特征时空变化大。在 1966-2015 期间,亚洲中部地区呈现整体湿润的趋势,但自 2003 年以来呈现较为明显的转干现象。亚洲中部的干旱存在较为明显的 16-64个月的周期性振荡且其干旱变化与 ENSO 的强度密切相关。其次,基于 1966 年 1 月至 2015 年 12 月期间 3 个月时间尺度的 SPEI,提出基于三维聚类算法的干旱事件识别和干旱特征描述体系,并应用于研究亚洲中部地区干旱的时空结构和动态特征。在经度-纬度-时间三维空间内,识别干旱事件并定量化分析干旱特征(如干旱严重程度、强度、干旱影响地区、质心和移动轨迹等)。结果表明, 亚洲中部地区干旱持续时间大多为 3 至 5 个月,其移动轨迹具有明显的东西方向性。长期干旱事件具有多峰特征。此外, 亚洲中部地区易发生春旱和夏旱。第三,提出了新的遥感降水产品误差成分评估体系,并将其应用于评估广泛使用的八种卫星降水反演产品(SPE),为干旱监测和其他应用提供科学的数据选择依据。为比较研究不同算法中的地面站点校正算法的有效性,八种产品同时包含了纯卫星降水反演产品和经过地面站点校正的卫星降水反演产品。结果表明,经过地面站点校正的卫星降水反演产品 GSMaP_Gauge 数据集在中亚具有最佳的表现,是遥感干旱监测的首选降水数据。其他数据集存在不同误差,表现相对较差,如对下垫面冻土和大型水体的高敏感性、显著的高程依赖误差等。最后, 在考虑降水与植被响应之间的滞后时间、最优权重以及干旱严重性等级重分类等方面的基础上,利用 GSMaP_Gauge 降水数据集以及 MODIS 提供的地表温度和归一化差值植被指数数据集,构建农业干旱遥感监测指数-最优比例干旱状态指数(OSDCI)。结果发现, OSDCI在描述农业干旱的空间格局和时间变化方面表现较好。 OSDCI 为基于遥感的高分辨率农业干旱监测提供了有效的监测工具。
Other AbstractArid Central Asia (ACA) is one region with semi-arid and arid climate and severewater resources deficit. Under the context of global warming and intense humanactivities, the frequent droughts lead to significant impacts on agricultural production,animal husbandry, water resources, economies and human health. Droughtidentification and characterization are particularly important to investigate the temporalvariations and spatial patterns of drought at different spatiotemporal scales and todevelop appropriate prevention and mitigation plans for drought disasters. It is also thecrucial method to understand drought events and their characteristics. However, thedrought characteristics are unclear and there are few studies on drought identificationand characterization in arid Central Asia due to both the difficulties for droughtmonitoring resulting from the varied and complex features and the limited datasets tosupport drought monitoring.Based on the above background, this dissertation aims to study the droughtcondition in arid Central Asia by considering two aspects: one is the droughtidentification and characterization based on the Standardized PrecipitationEvapotranspiration Index (SPEI) drought index and Climate Research Unit (CRU)datasets using both the traditional run theory method and the new-developed 3-dimensional clustering algorithm. The other one focuses on the application of remotesensing for agricultural drought monitoring by using a constraint optimal weightedmethod and minimum error iterative method based on multiple remote sensing variables.The main results of this dissertation are generally concluded as follows:Firstly, drought events, as well as their frequency, duration, severity, intensity andpreferred season, are studied by using the Run theory and SPEI drought index at 3-month, 6-month, and 12-month timescales. The Principle Components Analysis (PCA)and the Varimax rotation method were used to identify the sub-regional drought patterns,the Sen’s slope and the Modified Mann-Kendall method (MMK) were adopted to study the drought trend, while the wavelet analysis was applied to investigate the droughtperiodicity and the possible links between drought variation and large-scale climatepatterns. The results showed that the drought characteristics in ACA vary greatly. ACAhad an overall wetting trend during 1966-2015, but there is a significant drying trendsince 2003. The droughts in ACA had obvious 16-64-month periodical oscillation.Drought changes have a high association with the strength of ENSO.Secondly, an improved 3-dimensional clustering algorithm was promoted andapplied to study the space-time structure and characteristics of drought in ACA basedon the 3-month SPEI from January 1966 to December 2015. Within the longitudelatitude-time space, drought events were identified and their dynamic characteristics(e.g. severity, intensity, affected area, centroids and track path) were also quantified.Results showed that droughts in ACA tend to have a duration between 3 and 5 monthswith a preferred east-west/west-east trajectory. In addition, ACA tends to experiencedroughts in spring and summer.Thirdly, an error-component analysis evaluation system was refined and used toassess the most widely used eight satellite-based precipitation SPEs (SPE) to provide ascientific reference for drought monitoring and other applications in Central Asiacountries. Both satellite-only and gauge-corrected SPEs were selected to study thefunction of gauge-correction procedure in different algorithms. The overall conclusionbased on the results was that the gauge-corrected Global Satellite Mapping ofPrecipitation (GSMaP_Gauge) dataset had the best performance in this region which isthe best choice for drought monitoring while the other datasets suffered from diverseshortages like the high sensitivity of frozen land surface and large water bodies, theobvious elevation dependent accuracy and so on.Finally, an Optimal Scaled Drought Condition Index (OSDCI) was developedbased on the well evaluated GSMaP_Gauge precipitation dataset and the MODIS landsurface temperature and normalized difference vegetation index datasets by consideringthe varied lag time between precipitation and vegetation response for different landcover types, the objective and optimal selection of component weights and the revised severity classification scheme. It was found that OSDCI performed well in reflectingboth the spatial pattern and temporal variation of agricultural drought relative to the soilmoisture and multiple-scale SPEIs. OSDCI provides a good tool for drought monitoringbased on remote sensing at a fine resolution in arid Central Asia.
Subject Area地图学与地理信息系统
Document Type学位论文
First Author Affilication中国科学院新疆生态与地理研究所
Recommended Citation
GB/T 7714
郭浩. 近50年亚洲中部干旱特征研究[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
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[郭浩]'s Articles
Baidu academic
Similar articles in Baidu academic
[郭浩]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[郭浩]'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.