Effectiveness of hybrid ensemble machine learning models for landslide susceptibility analysis:Evidence from Shimla district of North-west Indian Himalayan region

The Indian Himalayan region is frequently experiencing climate change-induced landslides.Thus,landslide susceptibility assessment assumes greater significance for lessening the impact of a landslide hazard.This paper makes an attempt to assess landslide susceptibility in Shimla district of the northwest Indian Himalayan region.It examined the effectiveness of random forest(RF),multilayer perceptron(MLP),sequential minimal optimization regression(SMOreg)and bagging ensemble(B-RF,BSMOreg,B-MLP)models.A landslide inventory map comprising 1052 locations of past landslide occurrences was classified into training(70%)and testing(30%)datasets.The site-specific influencing factors were selected by employing a multicollinearity test.The relationship between past landslide occurrences and influencing factors was established using the frequency ratio method.The effectiveness of machine learning models was verified through performance assessors.The landslide susceptibility maps were validated by the area under the receiver operating characteristic curves(ROC-AUC),accuracy,precision,recall and F1-score.The key performance metrics and map validation demonstrated that the BRF model(correlation coefficient:0.988,mean absolute error:0.010,root mean square error:0.058,relative absolute error:2.964,ROC-AUC:0.947,accuracy:0.778,precision:0.819,recall:0.917 and F-1 score:0.865)outperformed the single classifiers and other bagging ensemble models for landslide susceptibility.The results show that the largest area was found under the very high susceptibility zone(33.87%),followed by the low(27.30%),high(20.68%)and moderate(18.16%)susceptibility zones.The factors,namely average annual rainfall,slope,lithology,soil texture and earthquake magnitude have been identified as the influencing factors for very high landslide susceptibility.Soil texture,lineament density and elevation have been attributed to high and moderate susceptibility.Thus,the study calls for devising suitable landslide mitigation measures in the study area.Structural measures,an immediate response system,community participation and coordination among stakeholders may help lessen the detrimental impact of landslides.The findings from this study could aid decision-makers in mitigating future catastrophes and devising suitable strategies in other geographical regions with similar geological characteristics.

Remote Sensing-based Spatiotemporal Distribution of Grassland Aboveground Biomass and Its Response to Climate Change in the Hindu Kush Himalayan Region

The grassland in the Hindu Kush Himalayan(HKH) region is one of the large st and most biodiverse mountain grassland types in the world,and its ecosystem service functions have profound impacts on the sustainable development of the HKH region.Monitoring the spatiotemporal distribution of grassland aboveground biomass(AGB) accurately and quantifying its response to climate change are indispensable sources of information for sustainably managing grassland ecosystems in the HKH region.In this study,a pure vegetation index model(PVIM) was applied to estimate the long-term dynamics of grassland AGB in the HKH region during 2000-2018.We further quantified the response of grassland AGB to climate change(temperature and precipitation) by partial correlation and variance partitioning analyses and then compared their differences with elevation.Our results demonstrated that the grassland AGB predicted by the PVIM had a good linear relationship with the ground sampling data.The grassland AGB distribution pattern showed a decreasing trend from east to west across the HKH region except in the southern Himalayas.From 2000 to 2018,the mean AGB of the HKH region increased at a rate of 1.57 g/(m~2·yr) and ranged from 252.9(2000) to 307.8 g/m~2(2018).AGB had a positive correlation with precipitation in more than 80% of the grassland,and temperature was positively correlated with AGB in approximately half of the region.The change in grassland AGB was more responsive to the cumulative effect of annual precipitation,while it was more sensitive to the change in temperature in the growing season;in addition,the influence of climate varied at different elevations.Moreover,compared with that of temperature,the contribution of precipitation to grassland AGB change was greater in approximately 60% of the grassland,but the differences in the contribution for each climate factor were small between the two temporal scales at elevations over 2000 m.An accurate assessment of the temporal and spatial distributions of grassland AGB and the quantification of its response to climate change are of great significance for grassland management and sustainable development in the HKH region.

Changes in extreme precipitation events over the Hindu Kush Himalayan region during 1961e2012

Based on a new multi-source dataset (GLDP-V1.0) recently developed in China Meteorological Administration, we employed precipitation indices including percentile-based indices of light (below the 50th percentile), moderate (between the 50th and 90th percentile), and intense (above the 90th percentile) precipitation, maximum 1-day, 3-day, and 5-day precipitation amounts (RX1DAY, RX3DAY, and RX5DAY, respectively), and consecutive wet and dry days (CWDs and CDDs) to analyze variations in extreme precipitation events in the Hindu Kush Himalayan (HKH) during 1961e2012. The main results are presented as follows. Firstly, there was a significant increase in the amount of light and moderate precipitation and number of associated days over various parts of India and northern Tibetan Plateau during 1961e2012; but the intensity of light precipitation decreased significantly in the Hindu Kush and central India, and the regional average intensity also decreased. Secondly, the amount and frequency of intense precipitation mostly increased significantly on the Tibetan Plateau, but there was a heterogeneous change over the remainder of the HKH, and regional average annual intense precipitation amount and frequency significantly increased over the HKH during 1961e2012. Thirdly, regional average RX1DAY, RX3DAY, and RX5DAYall showed significant upward trends during 1961e2012, and there was a significant increased tendency of consecutive wet-days in most parts of the study region; however, trends of consecutive dry-days were mostly opposite to those of consecutive wet-days, with regional averaged consecutive dry-days showing no noticeable trend.

Projected changes in mean and extreme climates over Hindu Kush Himalayan region by 21 CMIP5 models

Based on the outputs from 21 CMIP5 (Coupled Model Intercomparison Project phase 5) models, future changes in the mean temperature, precipitation and four climate extreme indices (annual maximum of daily maximum temperature (TXx), minimum of daily minimum temperature (TNn), annual total precipitation when the daily amount exceeds the 95th percentile of wet-day precipitation (R95p), and maximum consecutive 5-day precipitation (RX5day)) over Hindu Kush Himalayan (HKH) region are investigated under the greenhouse gas concentration pathways of RCP4.5 and RCP8.5. Two periods of the 21st century, 2036e2065 and 2066e2095, are selected, with the reference period is considered as 1976e2005. Results show general increase of the mean temperature, TXx and TNn under both scenarios, with the largest increases found during 2066e2095 under RCP8.5. Future precipitation is projected to increase over most part of HKH, except for the northwestern part. Intensification of the precipitation extremes is projected over the region. The uncertainties of mean temperature, TXx and TNn over the HKH1 subregions are the largest compared to the other three subregions and the overall HKH. Besides RX5day during 2036e2065 over HKH1, the uncertainties of R95p and RX5day tend to be larger following the increase of greenhouse gas concentrations. The multimodel ensemble medians of temperature and four extreme indices under RCP8.5 are projected to be larger than those under RCP4.5 in each of the subregions.

Downscaled climate change projections for the Hindu Kush Himalayan region using CORDEX South Asia regional climate models

This study assessed the regional climate models (RCMs) employed in the Coordinated Regional climate Downscaling Experiment (CORDEX) South Asia framework to investigate the qualitative aspects of future change in seasonal mean near surface air temperature and precipitation over the Hindu Kush Himalayan (HKH) region. These RCMs downscaled a subset of atmosphere ocean coupled global climate models (AOGCMs) in the Coupled Model Intercomparison Project phase 5 (CMIP5) to higher 50 km spatial resolution over a large domain covering South Asia for two representation concentration pathways (RCP4.5 and RCP8.5) future scenarios. The analysis specifically examined and evaluated multi-model and multi-scenario climate change projections over the hilly sub-regions within HKH for the near-future (2036e2065) and far-future (2066e2095) periods. The downscaled multi-RCMs provide relatively better confidence than their driving AOGCMs in projecting the magnitude of seasonal warming for the hilly sub-region within the Karakoram and northwestern Himalaya, with higher projected change of 5.4
C during winter than of 4.9
C during summer monsoon season by the end of 21st century under the high-end emissions (RCP8.5) scenario. There is less agreement among these RCMs on the magnitude of the projected warming over the other sub-regions within HKH for both seasons, particularly associated with higher RCM uncertainty for the hilly sub-region within the central Himalaya. The downscaled multi-RCMs show good consensus and low RCM uncertainty in projecting that the summer monsoon precipitation will intensify by about 22% in the hilly subregion within the southeastern Himalaya and Tibetan Plateau for the far-future period under the RCP8.5 scenario. There is low confidence in the projected changes in the summer monsoon and winter season precipitation over the central Himalaya and in the Karakoram and northwestern Himalaya due to poor consensus and moderate to high RCM uncertainty among the downscaled multi-RCMs. Finally, the RCM related uncertainty is found to be large for the projected changes in seasonal temperature and precipitation over the hilly sub-regions within HKH by the end of this century, suggesting that improving the regional processes and feedbacks in RCMs are essential for narrowing the uncertainty, and for providing more reliable regional climate change projections suitable for impact assessments in HKH region.

Invasive Alien Plants of Indian Himalayan Region—Diversity and Implication

The present study deals with comprehensive list of Invasive alien plants of Indian Himalayan Region with background information on family, habit and nativity. A total of 190 invasive alien species under 112 genera, belonging to 47 families have been recorded. Among these, the dicotyledons represent by 40 families, 95 genera and 170 species;monocotyledons represent by 7 families, 17 genera and 20 species. The analysis of invasive species reveals that 18 species have been introduced intentionally, while the remaining species established unintentionally through trade. In terms of nativity, amongst 13 geographic regions, the majority of invasive plants reported from American continent (73%). While in life form analysis, the herbs (148 species) are dominant, followed by shrubs (19 species), Grass (11 species), Trees (4 species), sedges and climber (3 species each). Most of the invasive species are annual habit (63%). Apart from these, 90 species (47%) are being used by locals for medicinal purposes. A better planning is needed for early detection to control and reporting of infestations of spread of new and naturalized weeds to be monitored.

Diversity, Distribution Pattern and Conservation Status of the Plants Used in Liver Diseases/Ailments in Indian Himalayan Region

In the Indian Himalayan Region, th studies focused on diversity of the plants used fo treating liver diseases/ailments have not been carried out so far. Therefore, the present attempt has been made to study the diversity, distribution pattern and conservation status of the plant species used fo treating liver diseases/ailments in that region. A tota of 138 species (35 species of trees, 22 shrubs and 8 herbs) belonging to 98 genera in 60 families hav been recorded. Amongst the families, Euphorbiacea (9 species), and altitudinal zone <1,800 m, (i.e., 11 species) are rich in species. Traditionally, variou plant parts, such as roots/rhizomes/tubers (46 species), leaves (31), whole plants (30), barks (15) fruits (13), seeds and unspecified parts (8 each), and inflorescence (1) are used for the treatment of live diseases/ailments. 34 species are native, 3 ar endemic and 15 near endemic. 7 species ar categorized as Critically Endangered (Betula utilis) Endangered (Podophyllum hexandrum, Ephedra gerardiana, and Nardostachys grandiflora) and Vulnerable (Bergenia ligulata, B. stracheyi, and Hedychium spicatum) using new IUCN criteria Available chemical composition of plant parts used fo the treatment of liver diseases/ailments have beengiven. Assessment of the populations of threatened species, development of an appropriate strategy, action plan for the conservation and sustainable utilization of such components of plant diversity are suggested.

Rhododendrons in Indian Himalayan Region: Diversity and Conservation

The genus Rhododendron of Indian Himalayan Region (IHR) has been enumerated in the present paper. A total of 87 species, 12 subspecies and 8 varieties of Rhododendrons recorded in IHR, among these 6 species and one subspecies are reported from Western Himalaya. The maximum concentration of 86% observed in Arunachal Pradesh (75 species). The species of Rhododendrons exhibit significant diversity in habit and broad range of distribution from the altitude of 800-6000 m. and the best range is observed in 3001-3500 m altitudes. In analysis revealed 20 taxa are endemic, 30 are rare, 24 are threatened / endangered, 3 are vulnerable and 47 taxa have to be assessed. The major threats to rhododendrons are deforestation and unsustainable extraction for firewood and incense by local people has been discussed.

A kinematics-uplift model for the HimalayanTibetan region

The Lhasa-Gangdise Terrane is taken as a representative mobile terrane during the Himalayan orogeny of the India- Eurasia continental collision, for which a corresponding kinematics-uplift model is set up. The parameterization of the model is ultimately constrained by the uplift history outlined by synthesized paleogeoglaphic studies with consideration of the following factors: (1 ) kinematic features of india-Eurasia plate convergence; (2) 3-D mass conservation during terrane deformations incorporating shortening, thickening, extension, uplift and ero sion; and (3) instantaneous vertical movement of lithospheric material under the control of isostasy. The model study involves the following four groups of uplift-relevant parameters: ① plate converging velocity and its variations with time; ② extent of lateral mass transfer; ③ crustal structure: and ④) surface erosion mode. The results of calculation of 144 models of different Parameter combinations have indicated the non-uniqueness of solution. Nevertheless, it is also proved that for a fixed kinematic mode of plate convergence there exists a unique best-fitting model which may reproduce the observed uplift history, implying the uniqueness of dynamic environment of two converging plates. Therefore, the uplift of the Himalayan-Tibetan region is mainly controlled by plate dynamics-kinematics and is a complicated geological process of far-reaching implications.

Historical temporal variation in precipitation over Western Himalayan Region：1857-2006

This study has examined the temporal variation in monthly, seasonal & annual precipitation over the Western Himalayan Region(WHR) and the influence of global teleconnections, like the North Atlantic Oscillation(NAO) and Southern Oscillation(SO) Indices on seasonal & annual precipitation. The Mann–Kendall non-parametric test is applied for trend detection and the Pettitt–Mann–Whitney test is used to detect possible shift. Maximum entropy spectral analysis is applied to find the periodicity in annual & seasonal precipitation. The study shows a non-significant decreasing trend in annual precipitation over WHR for the period 1857-2006. However, in seasonal precipitation, a significant decreasing trend is observed in monsoon and a significant increasing trend in post-monsoon season during the same period. The significant decrease in monsoon precipitation may be due to weakening of its teleconnection with NAO as well as SO Indices mainly during last three decades. It is observed that the probable change of year in annual & monsoon precipitation over WHR is 1979. The study also shows significant periodicities of 2.3-2.9 years and of 3.9-4.7 years in annual & seasonal precipitation over WHR.

Comprehending drivers of land use land cover change from 1999 to 2021 in the Pithoragarh District,Kumaon Himalaya,Uttarakhand,India

The Himalayan region has been experiencing stark impacts of climate change,demographic and livelihood pattern changes.The analysis of land use and land cover(LULC)change provides insights into the shifts in spatial and temporal patterns of landscape.These changes are the combined effects of anthropogenic and natural/climatic factors.The present study attempts to monitor and comprehend the main drivers behind LULC changes(1999-2021)in the Himalayan region of Pithoragarh district,Uttarakhand.Pithoragarh district is a border district,remotely located in the north-east region of Uttarakhand,India.The study draws upon primary and secondary data sources.A total of 400 household surveys and five group discussions from 38 villages were conducted randomly to understand the climate perception of the local community and the drivers of change.Satellite imagery,CRU(Climatic Research Unit)climate data and climate perception data from the field have been used to comprehensively comprehend,analyze,and discuss the trends and reasons for LULC change.GIS and remote sensing techniques were used to construct LULC maps.This multifaceted approach ensures comprehensive and corroborated information.Five classes were identified and formed viz-cultivation,barren,settlement,snow,and vegetation.Results show that vegetation and builtup have increased whereas cultivation,barren land,and snow cover have decreased.The study further aims to elucidate the causes behind LULC changes in the spatially heterogeneous region,distinguishing between those attributed to human activities,climate shifts,and the interconnected impacts of both.The study provides a comprehensive picture of the study area and delivers a targeted understanding of local drivers and their potential remedies by offering a foundation for formulating sustainable adaptation policies in the region.

西藏当雄县九子拉蛇绿混杂岩中超基性岩地球化学特征及地质意义

通过对九子拉区域出露的蛇绿混杂岩中超基性岩进行了岩石学、岩相学、地球化学特征等方面的研究,结果表明:九子拉地区超基性的岩石类型为辉橄岩,岩石具强烈蛇纹石化。主量元素方面表现出高Mg(MgO=43.27%45.46%)、高Si(SiO_(2)=43.93%46.52%)、贫Al(Al_(2)O_(3)=0.52%0.86%)、低铁Fe(FeOT=8.33%8.84%)的特征,经球粒陨石标准化的稀土元素配分模式表现为从轻稀土略富集趋势向右缓倾的特征,其(La/Yb)N为3.29.42,具较轻铕负异常(Eu/Eu^(*)=0.151.99)和较轻正铈异常(Ce/Ce^(*)=0.711.14)。微量元素明显富集大离子亲石元素K、Rb、P,相对富集Rb、Ta等元素,贫Hf、Sm、Yb元素,属典型的镁铁质超基性岩。样品中LA-ICP-MS锆石U-Pb最小年龄为119 Ma。结合区域构造背景,认为九子拉地区中部可能是一个早侏罗世开始俯冲、早白垩世闭合的具有独立演化历程的局限性洋盆,代表了特提斯洋向南俯冲诱导出的一系列藕断丝连的中生代弧间及弧后扩张盆地。

Snowmelt runoff prediction under changing climate in the Himalayan cryosphere: A case of Gilgit River Basin

There are serious concerns of rise in temperatures over snowy and glacierized Himalayan region that may eventually affect future river flows of Indus river system. It is therefore necessary to predict snow and glacier melt runoff to manage future water resource of Upper Indus Basin(UIB). The snowmelt runoff model(SRM) coupled with MODIS remote sensing data was employed in this study to predict daily discharges of Gilgit River in the Karakoram Range. The SRM was calibrated successfully and then simulation was made over four years i.e. 2007, 2008, 2009 and 2010 achieving coefficient of model efficiency of 0.96, 0.86, 0.9 and 0.94 respectively. The scenarios of precipitation and mean temperature developed from regional climate model PRECIS were used in SRM model to predict future flows of Gilgit River. The increase of 3 C in mean annual temperature by the end of 21 th century may result in increase of 35-40% in Gilgit River flows. The expected increase in the surface runoff from the snow and glacier melt demands better water conservation and management for irrigation and hydel-power generation in the Indus basin in future.

Impact of temperature and precipitation lapse rate on hydrological modelling over Himalayan Gandak River Basin

The hydrology of Himalayan region is influenced by temperature lapse rate(TLAPS)and precipitation lapse rate(PLAPS).Therefore,hydrological modeling considering TLAPS and PLAPS is crucial to manage the water resources in these terrains.In this research,Himalayan Gandak River basin is considered as the study area where TLAPS and PLAPS vary significantly due to high altitude of Himalayas.To assess the impact of TLAPS and PLAPS on water balance components,Soil Water Assessment Tool(SWAT)model was calibrated(2000-2007)and validated(2008-2014)on daily time step for three projects i.e.,Reference Project(RP),Snowmelt Project(SP)and distributed elevation band snowmelt project(SWAT-ETISM).The analysis discloses that SWAT-ETISM model(which has TLAPS and PLAPS parameters)outperforms the RP and the SP models in predicting streamflow with improved statistical indicators R2=0.88,NSE=0.84 and PBIAS=11.9.Furthermore,it was observed that SWAT-ETISM model comprehensively improved the streamflow statistics by improving the snow water equivalent and water balance components through the consideration of TLAPS and PLAPS values for the region.Hence,the proposed SWAT-ETISM model can be used for estimation of the water budget at the high-altitude and data scarce alpine Himalayan regions and worldwide,where PLAPS and TLAPS are substantial due to altitudinal variation.

环喜马拉雅地区三大英雄史诗题材比较研究——以《格萨尔》《罗摩衍那》《摩诃波罗多》为中心

诸多英雄史诗的题材有着“异中之同”的特点,其根本原因在于“人的一致性”。流传于环喜马拉雅地区的三大英雄史诗从故事情节和流传轨迹来看,相互无实质性影响,却有多种同一类型的题材。这些题材以“人的一致性”书写了环喜马拉雅地区民族的“英雄时代”,成为平行研究的故事基础。

喜马拉雅东构造结地震精定位及其区域应力场研究

喜马拉雅东构造结地处印欧大陆碰撞前缘,主要受喜马拉雅、拉萨、羌塘、川滇等地块和印度板块相互作用,区域构造变形强烈,是喜马拉雅造山带变形最强烈的地区之一,地震频发且主要呈条带状展布.为揭示该地区地震活动及发震机制、断裂现今运动状态和区域应力应变模式,本文以喜马拉雅东构造结及周缘地区为研究区,采用双差定位法对2008—2018年间65663个M≥1.0的地震事件进行重定位,应用CAP方法对2009—2021年间163个M≥3.5的地震事件进行震源机制解反演.在此基础上,收集研究区前人所得震源机制解共1156个,使用区域阻尼应力张量反演获得了中上地壳(0~35 km)区域应力场.研究结果显示,区内地震主要沿断裂展布,其中喜马拉雅东构造结、高原中部拉张裂谷、川滇地块和滇缅地块地震活动频繁.地震深度主要分布于5~25 km,川滇和滇缅地块内部地震相对于拉萨、羌塘地块的数量和优势深度有明显增大.不同类型的震源机制分布具有明显规律性,东构造结处各种机制类型地震频发;走滑型震源机制主要沿大型边界断裂分布;正断机制地震发生于川滇地块的西边界断裂;逆断地震发育于印欧大陆碰撞前缘.研究区主压应力轴水平方向从喜马拉雅、拉萨、羌塘、川滇、滇缅地块大致以东构造结为中心近顺时针旋转,且东构造结顶部、川滇地块北西部等地区呈现出强烈的局部不均匀性.

中国西藏南部喜马拉雅相的乐平统

西藏南部二叠系色龙群、曲布日嘎组等“喜马拉雅相”地层产有冈瓦纳区系的以 Spiriferella ra-jah,Taeniothaerus densipustulatus,Neospirifer(Neospirifer)kubeiensis和Retimarginifera xizangensis为典型代表的腕足动物群。其组成和演进层序与属于冈瓦纳大陆北缘的巴基斯坦盐岭的Wargal组上部和Chhidru组、克什米尔的Zewan组,印度斯区提Kuling页岩上部的Gungri组、尼泊尔西北部的Senja组和澳大利亚西部Hardman组的很接近;时代曾被归入瓜达鲁普世或乐平世早期等,分歧较大。近年来在盐岭等地发现这一冈瓦纳区系动物群明显高于乐平世下部牙形类 Clarkina dukouensis带,并与菊石Cyclolobus和有孔虫Colaniella动物群共生。由此推定整个色龙群或曲布组和曲布日嘎组都应属于乐平统,它与上覆三叠系以牙形类化石Hindeodus parvus和菊石Otoceras出现为界,代表冈瓦纳大陆北缘乐平世沉积的一个三级地层层序。层序以海进初期沉积的低水位体系域的曲布组石英砂岩为标志,其顶界则为比二叠系—三叠系界线略低的一个快速海进面。

中国喜马拉雅地区的主要蜜源植物研究

系统介绍了中国喜马拉雅地区， 包括西藏南部和东部、四川西南部、云南西部的蜜源植物资源概况， 主要蜜、粉源植物１１８ 种的开花期和利用价值， 为这一地区发展养蜂业， 振兴山区经济， 促进农、林业持续发展提供科学依据．

喜马拉雅琼嘉岗超大型伟晶岩锂矿的形成时代、源区特征及分异特征

喜马拉雅新生代淡色花岗岩带是近年来提出的与高度结晶分异、异地深成淡色花岗岩有关的稀有金属战略远景区,目前其金属组合以铍-铌-钽(-锡-钨)为主。秦克章等(2021a)报道了在高喜马拉雅带珠峰地区发现的琼嘉岗锂矿,是喜马拉雅首例具有工业价值的伟晶岩型锂矿。本次研究重点揭示喜马拉雅琼嘉岗伟晶岩型锂矿的成矿特征、形成时代和源区特征。琼嘉岗矿区矿石矿物主要为锂辉石、铌铁矿-铌锰矿、少量锡石和绿柱石,特征性长柱状锂辉石主要产于块体微斜长石+锂辉石带和分层细晶岩带内。琼嘉岗锂辉石伟晶岩各结构分带的K/Rb含量较为相似,锂含量从边部细粒钠长石带(~100×10^(-6))到分层细晶岩带(~1000×10^(-6)),再到块体微斜长石+锂辉石带(>3000×10^(-6))逐渐升高,而Cs含量逐渐降低。独居石和铌钽铁矿族矿物LA-ICPMS定年结果显示,琼嘉岗锂辉石伟晶岩形成于新喜马拉雅阶段早期(25~24Ma),与高喜马拉雅地区淡色花岗岩时代相近。矿物化学和独居石Nd同位素结果显示该稀有金属伟晶岩结晶于高度演化的花岗伟晶岩熔体,源区特征与高喜马拉雅结晶岩系一致。本研究所揭示的琼嘉岗成矿特征、形成时代和源区特征将为高喜马拉雅其它地区找寻大型花岗伟晶岩型锂矿提供重要借鉴意义。

基于地理探测器与SVM的冰湖溃决预测研究——以喜马拉雅山地区为例

冰湖溃决不仅对财产和基础设施具有破坏性,而且对当地居民也构成极大威胁。冰湖溃决的预测和风险评估对于预防和减轻灾害影响至关重要。文中提出了一个冰湖溃决的预测模型,强调选取容易获得的预测因子。以喜马拉雅山地区的48个冰湖为样本,使用地理探测器检测4个选定的预测因子:母冰川面积、冰舌坡度、冰湖面积和坝顶宽度。结果显示:冰舌坡度q值最大,为0.334 2。在交互作用检测器中,母冰川面积和冰舌坡度在交互作用后有最高的解释力,为0.684 4。这表明:与冰湖和冰碛坝相比,母冰川对冰湖状态的影响更大。在利用SVM(Support Vector Machine,支持向量机)构建的冰湖溃决预测模型中,验证集和测试集的准确率分别为83.33%和87.5%。研究为喜马拉雅地区未来的灾害管理提供了相应参考。