基于1961 -2017年5 -9月青藏高原99个地面气象观测站点的逐日降水资料, 选取10个极端降水指数, 采用线性倾向估计、 累积距平、 相关分析、 尺度分离等方法, 分析青藏高原暖湿季节极端降水的时空分布及变化特征。结果表明: 近57年来, 青藏高原暖湿季节降水强度、 1日最大降水量、 连续5日最大降水量显著增加, 进入21世纪后降水向强雨量雨日更多、 强度更强、 极值更大、 时间更集中的方向发展; 极端降水指数普遍存在3年、 4~8年、 10~11年、 20~30年以及更长时间尺度的周期变化, 准3年周期振荡对极端降水的贡献最大; 各极端降水指数之间联系密切, 中雨以上天数与暖湿季节降水总量的相关性最好; 降水总量、 强度、 强雨量雨日、 极值均由西向东、 由北向南增强增多, 降水强度、 大雨以上天数还随海拔高度的增加而显著减弱和减少, 最长连续有降水日数自北向南、 由低向高递增, 最长连续无降水日数由西向东递减; 东北及西南部极端降水事件增加最显著, 持续指数倾向率空间差异大, 其中最长连续无降水日数倾向率自西向东、 由高向低递增; 北大西洋多年代际振荡（the Atlantic Multidecadal Oscillation, AMO）和厄而尼诺南方涛动（El Ni?o-Southern Oscillation, ENSO）对青藏高原暖湿季节极端降水增多增强有一定影响。

为评估长序列降水时空变化及其对旱涝灾害的影响,采用趋势分析、累积距平分析等方法,选择降水集中指数(PCI)、降水异常指数(PAI)等指标分析了湖北省2000—2019年20 a间降水时空分布特征以及旱涝灾害发生状况。结果表明:近20 a湖北省累计年降水量增加了0.03%,降水多集中在4—8月份,雨热同期;2019年较2000年东部降水量增加;纬度决定了省内降水量的基数,总体呈现南部多于北部,经度影响降雨量变幅,表现为东西方向降水量空间分布的较大变异性;20 a间的PCI呈下降趋势,降水季节性和时段性特征较明显,PAI显示2005—2015年期间具有洪涝转干旱的趋势,2015年SPI值最大为1.425,出现了较严重洪涝;总体来看湖北省旱涝灾害呈现周期性波动特点,时间周期特征显著。研究成果对年际尺度上区域气候变化应对、旱涝灾害预警等方面具有重要意义。

<p>China&rsquo;s southwestern special terrain pattern as parallel arrangement between longitudinal towering mountains and deep valleys has significant effects on the differentiation of local natural environment and eco-geographical pattern in this region. The 1:50,000 Digital Elevation Model (DEM) data of Longitudinal Range-Gorge Region (LRGR), meteorological observation data from the station establishment to 2010, hydrological observation data, Normalized Difference Vegetation Index (NDVI) and Net Primary Productivity (NPP) products of MOD13 and MOD17 as well as 1:1,000,000 vegetation type data were used. Moisture indices including surface atmospheric vapor content, precipitation, aridity/humidity index, surface runoff, and temperature indices including average temperature, annual accumulated temperature, total solar radiation were selected. Based on ANUSPLIN spline function, GIS spatial analysis, wavelet analysis and landscape pattern analysis, regional differentiation characteristics and main-control factors of hydrothermal pattern, ecosystem structure and function in this region were analyzed to reveal the effects of terrain pattern on regional differentiation of eco-geographical elements. The results show that: influenced by terrain pattern, moisture, temperature and heat in LRGR have shown significant distribution characteristics as intermittent weft differences and continuous warp extension. Longitudinal mountains and valleys not only have a north-south corridor function and diffusion effect on the transfer of major surface materials and energy, but also have east-west barrier function and blocking effect. Special topographic pattern has important influences on vegetation landscape diversity and spatial pattern of ecosystem structure and function, which is the main-control factor on vegetation landscape diversity and spatial distribution of ecosystem. Wavelet variance analysis reflects the spatial anisotropy of environmental factors, NDVI and NPP, while wavelet consistency analysis reveals the control factors on spatial distribution of NDVI and NPP as well as the quantitative relationship with control degree. Special terrain pattern in LRGR is the major influencing factor on eco-geographical regional differentiation in this region. Under the combined effect of zonality and non-zonality laws with &ldquo;corridor-barrier&rdquo; function as the main characteristic, special spatial characteristics of eco-geographical regional system in LRGR is formed.</p>

以长江源区为研究对象,采用沱沱河站、直门达站实测径流系列,分析两站近60年来径流演变趋势,利用海洋尼诺指数分析ENSO事件的强度与时间特征,研究两站径流丰枯变化对ENSO事件的响应规律,从海-气耦合影响大气环流角度分析ENSO事件对径流的影响机制。研究结果表明,两站均呈现震荡上升的趋势,尤其在2000年以后径流呈现显著增加的趋势。分析ENSO事件发生期间长江源区径流丰枯变化响应规律可知,1960—2000年期间,暖事件发生年份,无论是当年还是次年,长江源区出现径流偏枯的概率较高;冷事件发生年份,两站均呈现出在当年偏枯、次年丰枯概率基本相当的态势。2000年以后,长江源区径流偏丰态势尤为显著。在1970年以前和2000年以后,长江源区径流与ENSO事件之间的时频结构具有一定程度的正相关性。ENSO事件发生后,引起纬向和经向大气环流异常,使得海洋向高原热量和水汽输送产生变化,从而影响青藏高原降水和径流。

横断山区是中国山洪灾害发育最集中、灾情最严重的区域之一，也是西南重大工程、美丽乡村建设的重点部署区。然而，现有研究对横断山区的山洪灾害发育特征与形成模式尚缺乏系统梳理和总结，制约了西南山区防洪理论和技术体系的发展完善。本文聚焦横断山区山洪灾害形成、运动、成灾的物理过程，提炼了横断山区山洪的主导孕灾条件（多重湿润季风环流条件、高位能地形地貌条件与高异质性下垫面条件）和灾害发育特征（时空分异、能量集中、链生性与群发性），从灾害全过程的视角提出了横断山区山洪灾害的三大形成模式：高水力坡降主导的生—水—土耦合的山地产汇流模式、山洪运动过程的强水沙耦合运动模式、承灾体较高暴露度导致灾害损失严重的成灾模式。最后，基于现有研究存在的问题提出了横断山区山洪灾害研究面临的四大未来科学挑战，以期推动横断山区乃至全国的山洪理论体系建设、区域山洪灾害风险精准评估与精细化预警预报等防灾减灾基础研究工作。

The Hengduan Mountains Region is one of the regions with the most densely distributed and severe flash flood disasters in China. It is also the key area for major engineering projects and beautiful countryside construction in southwest China. However, previous studies have not systematically summarized the development characteristics and formation modes of flash flood disasters in this region, which restricts the development and establishment of flood control theory and technical system in the southwestern mountains. This paper focuses on the physical processes of generation, movement and hazard-formation of flash flood disasters in the Hengduan Mountains. It clarifies the dominant disaster-inducing conditions (multiple humid monsoon circulation conditions, high potential energy conditions and high heterogenous underlying surface conditions) and disaster development characteristics (high spatio-temporal heterogeneity, highly concentrated energy, disaster chain and clustered occurrence) of flash floods in the Hengduan Mountains. Based on the entire processes of flash flood disasters, three major formation modes are summarized: the runoff generation of vegetation-hydrology-soil coupling dominated by high hydraulic gradient in mountainous areas, strong runoff-sediment coupled movement, and serious disaster losses due to high exposure of disaster bearing objects. Finally, based on the issues in previous research, four future research challenges for flash flood disasters in the Hengduan Mountains are proposed. Our work contributes to the development of disaster prevention and reduction research, including basic theoretical system, precise risk assessment of regional disasters, and fine early warning and forecasting of flash floods.

基于高空间分辨率0.25°的ERA-Interim再分析资料、TRMM 3B43 Version7数据、气象站点实测数据等多源数据，本文采用一种新的流域边界水汽通量概化和提取方法，揭示了夏季怒江流域水汽输送多支特征，并分析了其对降水时空分异的影响。研究表明，在高黎贡山南部、北部，伯舒拉岭北部及念青唐古拉山中部，有4支区域性水汽输送高值区，多年平均输送通量分别达102.6 kg/(m·s)、66.3 kg/(m·s)、39.7 kg/(m·s)和41.3 kg/(m·s)。多支水汽输送不仅深刻影响流域水汽输送格局，而且对降水时空分异也有不同程度影响。年际变化上，中下游横断山区水汽输送对降水的影响较小，上游青藏高原区影响较大，尤其以那曲—比如—索县一带影响最为显著。空间分布上，流域降水与水汽输送通量呈显著正相关，受多支水汽输送影响形成多个区域性多雨带。

Water vapor transport is a key factor in the transformation of atmospheric water resources to terrestrial water resources. In order to reveal the characteristics of water vapor transport and its influence on precipitation over the Nujiang River basin, high spatial resolution (0.25°) ERA-Interim reanalysis data, TRMM 3B43 Version7 data, meteorological station measured data and SRTM Version4.1 DEM data were used, and a new method to generalize and extract water vapor transport data along the watershed boundary was adopted. Through analysis and research, the multiple branches water vapor transports over the Nujiang River Basin in summer were detected, and their effects on the spatiotemporal distribution of precipitation were analyzed. The research shows that there are four regional high-value areas of water vapor transport on the western boundary of the watershed, which are located in the southern and northern parts of Gaoligong Mountain, the northern part of Boshula Mountain, and the middle of Nyenchen Tanglha Mountains, with the annual average water vapor flux being 102.6 kg/(m·s), 66.3 kg/(m·s), 39.7 kg/(m·s), and 41.3 kg/(m·s), respectively. Multiple branches water vapor transport affects not only the water vapor transport in different areas of the basin, but also the spatiotemporal characteristics of precipitation. In terms of interannual variation, the influence of water vapor transport on precipitation is weaker in the middle and downstream area (Hengduan Mountains), while it is greater in the upstream area (Qinghai-Tibet Plateau), especially in the Nagqu-Biru-Suoxian area. In terms of spatial distribution, annual precipitation in the basin is significantly positively correlated with water vapor flux and negatively correlated with water vapor flux divergence. Because of multiple branches of water vapor transport, four regional rainy areas are formed by the west side of the basin.