• <tr id="yyy80"></tr>
  • <sup id="yyy80"></sup>
  • <tfoot id="yyy80"><noscript id="yyy80"></noscript></tfoot>
  • 99热精品在线国产_美女午夜性视频免费_国产精品国产高清国产av_av欧美777_自拍偷自拍亚洲精品老妇_亚洲熟女精品中文字幕_www日本黄色视频网_国产精品野战在线观看 ?

    Thickness estimation of the Longbasaba Glacier:methods and application

    2020-03-29 08:06:52GuangLiHeJunFengWeiXinWang
    Sciences in Cold and Arid Regions 2020年6期

    GuangLi He,JunFeng Wei*,Xin Wang,2

    1. School of Resource Environment and Safety Engineering, Hunan University of Science and Technology, Xiangtan,Hunan 411100,China

    2. State Key Laboratory of Cryospheric Science, Northwest Institute of Ecology and Environmental Resources, Chinese Academy of Sciences,Lanzhou,Gansu 730000,China

    ABSTRACT A total of 71,177 glaciers exist on the Qinghai-Tibet Plateau,according to the Randolph Glacier Inventory (RGI 6.0).De‐spite their large number,glacier ice thickness data are relatively scarce.This study utilizes digital elevation model data and ground-penetrating radar thickness measurements to estimate the distribution and variation of ice thickness of the Longba‐saba Glacier using Glacier bed Topography (GlabTop), a full-width expansion model, and the Huss and Farinotti (HF)model. Results show that the average absolute deviations of GlabTop, the full-width expansion model, and the HF model are 9.8, 15.5, and 10.9 m, respectively, indicating that GlabTop performs the best in simulating glacier thickness distribu‐tion.During 1980?2015,the Longbasaba Glacier thinned by an average of 7.9±1.3 m or 0.23±0.04 m/a,and its ice volume shrunk by 0.28±0.04 km3 with an average reduction rate of 0.0081±0.0001 km3/a.In the investigation period,the area and volume of Longbasaba Lake expanded at rates of 0.12±0.01 km2/a and 0.0132±0.0018 km3/a,respectively.This proglacial lake could potentially extend up to 5,000 m from the lake dam.

    Keywords: Longbasaba Glacier; ice thickness distribution; ice volume change; glacial lake retention role; glacial lake expansion

    1 Introduction

    In the context of global warming, glaciers around the world have generally continued to melt over re‐cent decades (Lemkeet al., 2007). Such morphologi‐cal change, which is characterized by reduced glacier area and thickness (Beameret al., 2017; Mayeret al.,2018; Van Tielet al., 2020), has important implica‐tions regarding local/regional water resources(Immer‐zeelet al.,2010;Immerzeel and Bierkens,2012;Port‐neret al., 2019). Data on glacier thickness distribu‐tion and change are useful in various respects: (i) as basic parameters of glaciological research and applica‐tion, (ii) as essential input parameters for establishing glacier hydrological models, dynamic models, and di‐saster simulation models (Huggelet al., 2003), and(iii) as important indices for both evaluating the influ‐ence of glacier change on surface runoff and formulat‐ing disaster prevention/mitigation measures (Zhanget al., 2012). Owing to limited access and economic costs associated with field-based glacier investigation,it is difficult to directly measure the distribution of glacier thickness and estimate the total glacier volume at the regional scale. Comprehensive, accurate data of glacier ice thickness and ice volume are extremely rare. For glaciers that are difficult to access for direct measurement or that have a large spatial scale, one reasonably effective approach is to estimate glacier thickness distribution based on glacier development characteristics and topographic factors for determin‐ing glacier volume. Many researchers have developed various ice thickness estimation methods based on surface characteristics (Farinottiet al., 2017) that can be used to evaluate glacier ice thickness distribution and ice storage indirectly.

    Among the many existing methods for estimating glacier ice thickness distribution,the empirical formu‐la method is the simplest and most practicable, and it has been used widely in recent years (Bahret al.,2015). According to measured data of typical glacier thickness, the empirical formula method uses statisti‐cal methods to obtain a relationship for evaluating the ice thickness distribution and ice volume based on easily obtained glacier parameters such as area and length. This approach primarily explores the relation‐ship between the area and volume of a glacier (Chen and Ohmura, 1990; Bahret al., 1997) and also in‐cludes other characteristics such as glacier length or surface slope (Lüthi, 2009; Radi? and Hock, 2011;Grinsted, 2013). However, the empirical formula method can be used to estimate only the mean thick‐ness and total volume of a glacier.The method adopt‐ed for the estimation of ice thickness distribution usu‐ally depends on various theoretical considerations.Some studies have highlighted that, for an ideal gla‐cier with infinite width, ice thickness can be assessed from the surface slope (Nye, 1952). Subsequently,Nye (1965)extended the scope of consideration to the ideal shape of a valley glacier,and Liet al.(2012)fur‐ther considered the influence of glacier edge side re‐sistance. Haeberli and Heozle (1995) were the first to suggest that glacier thickness could be estimated from the range of glacier elevations. Subsequently, a series of glacier thickness studies were conducted by glacier bed topography (referred to as the GlabTop model, or simply GalbTop hereafter), a modified version named GlabTop2, and the latest version called Glab‐Top2_IITB (Linsbaueret al., 2012; Freyet al., 2014;Pandit and Ramsankaran, 2020). The GlabTop is a model that evaluates the spatial distribution of ice thickness by estimating glacier depth at several points on the so-called glacier branch line. The GlabTop2 model randomly selects digital elevation model (DEM)units in the glaciated area. The GlabTop2_IITB is an independent implementation of the GlabTop2 model and is fully automated.

    Budd and Allison (1975) and Rasmussen (1988)applied the methods of mass conservation and ice flow dynamics, which were further developed for ice thickness analyses (Fastooket al., 1995; Farinottiet al.,2009a).Huss and Farinotti(2012)continued to ex‐tend the ice flow dynamics method and were the first to successively present a comprehensive ice thickness distribution of a single glacier.Alternative approaches based on rigorous inverse models often focus on infer‐ring basal slipperiness based on bedrock topography(Gudmundssonet al., 2001; Thorsteinssonet al.,2003; Pralong and Gudmundsson, 2011; Mosbeuxet al., 2016). In recent years, the development of meth‐ods for estimating glacier ice thickness distribution based on surface characteristics has increased rapidly.Some techniques need additional parameters,e.g.,sur‐face velocity and mass balance (Morlighemet al.,2011; McNabbet al., 2012; Clarkeet al., 2013; Fari‐nottiet al.,2013;Gantayatet al.,2014;Huss and Fari‐notti, 2014; Brinkerhoffet al., 2016). Other methods adopt an iterative approach using complex forward models of ice flow (Michelet al., 2013; Van Peltet al., 2013; Michelet al., 2014) or a nonphysical algo‐rithm based on neural network methods (Clarkeet al.,2009; Haqet al., 2014). Some models have also achieved good results in the estimation of ice thick‐ness.For instance,the Brinkerhoff-v2 model(Brinker‐hoffet al., 2016), widely considered the best, can achieve an estimate of glacier thickness with an aver‐age deviation of 3%±27% (Farinottiet al., 2017).Among the various automated methods that can deal with a large number of glaciers, the Huss model is the best, with an average deviation of -14%±35% (Huss and Farinotti,2012;Farinottiet al.,2017).

    The input data for an ice thickness distribution model include at least some information on the gla‐cier or ice cap profile and an ice surface DEM. Such models require additional information according to specific conditions, including the surface mass bal‐ance(SMB),ice thickness change rate(?h/?t),and sur‐face velocity. The contours and DEM of a glacier can be acquired from satellite images at a given time. In contrast, the SMB, ?h/?t, and surface velocity of the glacier usually refer to the multiyear averages of an epoch,which should be as close as possible to the cor‐responding DEM to ensure temporal consistency be‐tween datasets. It is difficult to estimate the internal surface velocity,however.

    Despite a large number of methods available for the assessment of glacier ice thickness, there has been no analysis of the applicability of any of these methods to a glacier on the Qinghai-Tibet Plateau.The remote location and harsh environment of the Qinghai-Tibet Plateau make it difficult to obtain gla‐cier thickness and ice volume data directly. Currently,glacier ice volume data are relatively scarce, despite a large number of glaciers on the Qinghai-Tibet Plateau.The Longbasaba Glacier, in direct contact with an gla‐cial lake, has a huge scale and many branches.There‐fore, the various ice flow directions and the shear stress on the branch lines are not the same as those on the main flowline,and it is inappropriate to directly es‐timate ice thickness utilizing the glacier height differ‐ence.Hence,this study estimated the ice thickness dis‐tribution of the Longbasaba Glacier using the Glab‐Top model, the full-width expansion model, and the HF-model.The model results were compared with the thickness measurements obtained simultaneously from ground-penetrating radar (GPR) and differential glob‐al positioning system (GPS) to validate their applica‐bility. Finally, the thickness distribution of the Long‐basaba Glacier in 1980,2000,2010,and 2015 estimat‐ed using the most suitable model was presented and used to investigate the variation of ice volume and the volume of the glacial lake.

    2 Study site

    The Longbasaba Glacier (27°56.67′N, 88°04.21′E)is within the Pengqu River Basin on the northern slope of the Himalayas.The Longbasaba Glacier,with a length of 9 km and an area of approximately 31 km2in 2018(Liuet al.,2020),has experienced rap‐id retreat since the 1970s (Wanget al., 2008; Yaoet al., 2012). Ice serac, supraglacial lakes, and ice cre‐vasses are well-developed features on the glacier sur‐face (Figure 1). The lower part of the ice cliff at the end of the glacier is recessed into the glacier. During the field investigation,the collapse of the ice cliff was observed frequently. Thus, it can be inferred that the thermal karst process plays an essential role in the ero‐sion at the end of the glacier.The glacier terminal mo‐raine lake (Longbasaba Lake) lies at the elevation of 5,500 m and has a length (N ?S) of 2,210 m and a width (E?W) of 685 m. This strip-shaped lake is the source of the second tributary of the Pengqu River Ba‐sin.The water supplied to the Longbasaba Lake is de‐rived mainly from glacier meltwater. In summer,many ice bodies with different sizes break away from the mother glacier and float on the lake.According to a field survey conducted using an echo sounder with a GPS receiver in 2009(Yaoet al.,2012),the water lev‐el of Longbasaba Lake is 5,499 m a.s.l..The lake's av‐erage depth and maximum depth are 48 and 102 m,re‐spectively, and its volume is 0.064 km3. The surface moraine of the glacial lake dam is composed mainly of medium?coarse granite.An artificial outlet was ex‐cavated at the southwest corner of the dam to drain the lake,forming many small ponds at the depressions in front of the moraine ridge. Based on soil tempera‐ture, water, and heat flux data observed at different depths of the dam during 2012?2016, it has been es‐tablished that the permafrost layer of the dam is dete‐riorating, compromising its stability (Wanget al.,2018). Consequently, Longbasaba Lake is considered a high-risk moraine lake, and the probability of col‐lapse is classified as"very high"(Wanget al.,2008).

    Figure 1 Location and characteristics of the study area:(a)glacier location;(b)contact of the end of the Longbasaba Glacier with the glacial lake,ice floes show a small amount of newly collapse;(c)profiles of GPR survey conducted on the Longbasaba Glacier in July 2009 with Landsat 5 image from December 30,2010,as the background

    3 Data and methods

    3.1 Data

    In this study, a topographic map and three Land‐sat images were used to extract boundary informa‐tion of the Longbasaba Glacier in different periods.In addition, terrain related parameters such as sur‐face slope (needed for ice thickness and ice volume estimation)were extracted for the corresponding pe‐riods from DEM data,e.g., the topographic map and the satellite-derived Shuttle Radar Topography Mission (SRTM), the Advanced Land Observing Satellite (ALOS), and the TerraSAR-X add-on for digital elevation measurements (TanDEM-X) data(Table 1).

    Table 1 Overview of the satellite imagery and data sources used in the study

    The Landsat images were obtained from the glob‐al remote sensing image database (https://earthexplor‐er.usgs.gov/) of the United States Geological Survey(USGS). Such images, which have 30-m spatial reso‐lution and plane positioning error better than one pix‐el (Bolchet al., 2010; Guoet al., 2013), are used widely to extract glacier area and related attributes(Guoet al.,2015).In this study,based on a topograph‐ic map from 1980 and Landsat images obtained in 2000, 2010, and 2015, the boundary information of the glacier/glacial lake in corresponding periods was obtained manually. The streamline information of the glacier in the four periods was obtained using an auto‐matic extraction method(Yaoet al.,2015).

    The topographic map was produced by the China Military Geodetic Service based on aerial photo‐graphs taken in October 1980. Contours were digi‐tized manually and then converted into a raster DEM(TOPO DEM) with 30-m grid pixels using the Thies‐sen polygon method (Shangguanet al., 2010; Weiet al.,2015;Zhanget al.,2016).Wuet al.(2018)report‐ed that the vertical accuracy of TOPO DEM on gla‐ciers is better than 9 m. The 2000 SRTM1 Arc-Sec‐ond Global data with 30-m resolution were retrieved from the International Scientific Data Mirroring Web‐site (http://datamirror.csdb.cn) of the Chinese Acade‐my of Sciences Computer Network Information Cen‐ter. The vertical difference in the mountain area, as calculated based on the root mean square error of the GPS data,was approximately 12 m (Khasanov,2020).The ALOS/PRISM Level 1B1 stereo image pair DEM was selected to extract the 2010 glacier elevation(http://218.248.0.130/internet/servlet/LoginServlet) with 30-m resolution. At last, the 2015 DEM data of the Longbasaba Glacier were obtained using the direct TanDEM-X image data product provided by the Ger‐man space agency. It is worth noting that the vertical accuracy of the ALOS and TanDEM-X data is 6 to 8 m in steep terrain(Rastner and Paul,2020).

    In this study, field measurements of the thickness of the Longbasaba Glacier obtained in 2009 were used to calibrate the parameters and to test the accura‐cy of the glacier thickness estimation. The procedure is as follows. (1) Obtain thickness measurements or estimates from GPR and model simulation at GPS points or sampling points based on DEM data; (2)find the differences between the two data and calcu‐late the average deviation (AD); (3) adjust parameters such asf,m,n,A,C,fdebris, and db?/dz0and iterate (1)and (2) until the AD value is smallest.The parameters are then deemed as optimal and used for estimating the glacier thickness at this time, and the AD value is considered as the accuracy.

    The thickness measurements were obtained using a pulse EKKOTM100 GPR (Sensors & Software Inc.,Canada) at 237 points along a line shown in Figure 1c. The thickness at each measuring point was deter‐mined using the Reflexw GPR image processing soft‐ware with a relative error of 1.18%, which is within the required range of glaciological accuracy (Sunet al.,2003).The GPR antenna was operated at 100 MHz at a spacing of 2.5 m and a step length of 0.5 m.The measured wave velocity was 0.169 m/ns. During the GPR survey, a GPS receiver was used to acquire three-dimensional coordinates of each measuring point.

    Yaoet al.(2012) used a SyQWest HelboxTMpor‐table high-resolution shallow-water echo sounder to conduct a wide range of on-site water depth measure‐ments on the Longbasaba Lake and collected a total of 35,558 measurement points. And the position of each water depth measurement was acquired by an eTrex Vista GPS receiver with a horizontal accuracy of about 3?6 m and an average error of 1?2 m in so‐nar depth data.

    3.2 Glacier thickness and volume estimation

    Owing to the limited availability of critical param‐eters in the Longbasaba Glacier area,this study select‐ed three methods to estimate glacier thickness: Glab‐Top,the full-width expansion model,and the HF-model.The accuracy of each method was evaluated based on measured data of ice thickness.

    (1)GlabTop:GlabTop assumes that the glacier is a U-shaped ideal plastic body. When in a stable state,the maximum thickness of the ice body can be evalu‐ated based on the glacier surface slope and basal shear stress (Paterson and Cuffey, 1994). It is generally be‐lieved that the maximum thickness of a glacier occurs at the mid-streamline; therefore, glacier thickness (hf)at the streamline can be expressed as follows:

    whereτrepresents the average basal shear stress along the mid-streamline kPa;ρ, ice density;g, gravi‐tational acceleration; andα, the average surface slope angle at the glacier mid-streamline, which can be cal‐culated from the DEM. For alpine glaciers, the mor‐phological factorfwas calibrated by the measured ice thickness data to be 0.8. The basal shear stress of the glacier was determined based on the glacier height dif‐ference(ΔH)(Haeberli and H?lzle,1995):

    Based on ALOS data, the elevation difference of the Longbasaba Glacier in 2010 was 1,953.88 m;there‐fore, the basal shear stress was 150 kPa. The average glacier surface slope was estimated based on the mean slope within a 400-m length of the mid-streamline(Paterson and Cuffey,1994).

    (2)Full-width expansion model:The full-width expansion model represents a development of Glab‐Top.In this method,the glacier shape factor is consid‐ered to have a significant influence on the estimation of maximum glacier thickness, and that fixed parame‐ters should not be used in the calculation for different glaciers (Liet al., 2012). Based on statistical analysis,Nye (1965) suggested that the shape factor is a func‐tion of the glacier's width and thickness. However, Liet al.(2012) assumed a parabolic cross-section and expressed the glacier shape factor (f) using the fol‐lowing formula:

    whereγis the ratio of the glacier's half-widthwto its thicknessh, andmis a constant (0.9). Therefore, the above equation can be transformed into the following(Liet al.,2012):

    whereHiis the thickness of the glacier in an ideal state,which can be calculated based on Equation(1).

    (3)HF-model:According to the basic principles of glacier mass transformation and ice flow dynamics(Glen, 1955), the HF-model considers that the thick‐ness at any point on a streamline of the glacier can be determined as follows:

    whereCis the terrain correction coefficient;A,the ve‐locity coefficient;n, the flow law index, which usual‐ly fluctuates between 1.5 and 4.2 (generally, taken as 3.0);the mean ice volume flux;the flow line along the glacier;ρ, the density of ice (kg/m); andg,gravitational acceleration (m/s2). The ice flux at any point on a streamline of a glacier can be computed based on the conservation of glacial mass and the gra‐dient of the glacial mass balance as follows (Farinottiet al.,2009a):

    wherefdebrisis the influence factor of the epidermis(generally, taken as 0.5) (Husset al., 2007). For each glacier, the apparent mass balance gradientsandand the elevationzwere used to calculate the apparent mass balance distribution. The apparent mass balancewas derived by accumulating the same at each grid unitiin the glacier area.in this ar‐ea is conducive to the calculation of ice volume flux.The normalized ice volume flux and the local glacier width relevant for ice discharge were used to obtain the mean specific ice flow

    Besides, the apparent mass balance gradientsandcould be calculated using the Huss method(Huss and Farinotti,2012):

    whereC0is the continentality index, which is intro‐duced to set the reference apparent mass balance gra‐dientat 5.5×10?3per year (Husset al., 2007)according to the local glacial conditions. Based on the glacial cell altitudeELAlatand the median altitudeELAg,C0can be obtained as follows:

    wherefcont, equal to 2,400 m, is a constant parameter calibrated by the difference of the mass balance gradi‐ent observed in mainland glaciers and marine glaciers(Haeberliet al.,2008).

    Based on the measured ice thickness data, the ice thickness estimates using GlabTop, the full-width ex‐pansion model, and the HF-model were analyzed and compared to select the optimal method for estimating the thickness of the Longbasaba Glacier. The glacier was divided into different elevation zones based on 50-m elevation intervals. The thickness at the middle point of the different elevation zones of the glacier mid-streamline was evaluated using the above three methods. Additionally, the glacier thickness at any point of the mid-streamline was estimated using the inverse distance weighting method. Based on the zero ice thickness assumption at the glacier boundary and ridge line, and glacier streamline thickness distribu‐tion,the Kruger spatial interpolation method was used to interpolate the thickness at any point of a glacier to obtain the characteristics of glacier thickness distribu‐tion. Notably, topographic factors such as glacier sur‐face slope at selected periods were extracted from a 1980 topographic map, 2000 SRTM data, 2010 ALOS data,and 2015 TanDEM-X data.Eventually,the thick‐ness distributions were obtained for estimating the glacier's volume in the different periods and analyzing its variation characteristics. Specifically, the ice vol‐ume (Vg) was estimated based on the area (Sg)and the average glacier thicknessby using the following formula:

    3.3 Estimation of glacial lake water volume

    The topography of the Longbasaba Lake/Glacier bed was constructed to calculate the amount of lake water during different periods. First, the depth data collected in 2009 was utilized to establish the shape of the lake basin (Yaoet al., 2012), and the ice thick‐ness and topographic maps of 1980 were used to ob‐tain topographic data of the glacier bed. Considering the steep slope at the glacier end, a vertical glacier front,or a straight"line"in a section view(refer to Fig‐ure 6), was assumed when merging the bed terrains of the lake and glacier. The lake water volume in each year,Vl,was estimated using the following equation:

    where,Nlis the pixel number within the lake bound‐ary in each year;the depth of the individual pixel,which can be calculated by comparing the water level and the basin morphology; andS, the lake area in the corresponding year.

    3.4 Uncertainty analysis

    The thickness estimates from the three methods were compared with the field measurements by GPR.Each method's total error was quantified by the aver‐age difference between the estimated and the mea‐sured thicknesses (hereafter, the average deviation).The average deviation expressed as a percentage of each glacier's mean measured flow line thickness was defined as the ice thickness uncertainty (Liet al.,2012).

    The geolocation errors of the pixels on the glacier/lake boundaries, generated through a careful manual approach, can be controlled within a pixel. By com‐paring the outline positions from the manual delinea‐tion and GPS measurements, Guo and others (2015)suggested that the accuracies of manual outlines were±11 m and±30 m for clean ice and debris-covered gla‐ciers, respectively. In this study, the position accura‐cy of glacier boundary was affected by the debris cover on the glacier tongue but without the glacier terminus, as it was easy to distinguish the glacier from its proglacial lake exactly. The accuracies of the generated area were defined by the buffer along the glacier/lake perimeters with the size of half of the outline position error, and equal to 0.5 pixels multiplied by the pixel number within the perimeters and the spatial resolution of the images (Bolchet al.,2010; Guoet al., 2015). The average area uncertain‐ties of the glacier and lake were found to be ±8%and±3%,respectively.

    According to the ice thickness calculation formu‐la, the uncertainty in ice volume estimation depends on ice thickness and glacier area uncertainty and was found to be about 15.2%. Since the basin topography was derived from ice thickness data and sonar detec‐tion data,the uncertainty of lake water volume estima‐tion results from two aspects. The first aspect of the uncertainty is related to the estimation of the lake wa‐ter volume before the fusion line and mainly depends on the accuracy of the sonar sounding.It can be evalu‐ated as the product of the lake area error and the aver‐age lake depth. The uncertainty of the lake water vol‐ume after the fusion line (basin of 1980) should con‐sider not only the accuracy of sonar sounding but also the uncertainty of the ice thickness estimation and can be evaluated by the following formula:

    where,?vis the uncertainty of the glacial lake volume;and ?s, the uncertainty of the lake area;is the aver‐age water depth of the lake;is the uncertainty of the average water depth;Sis the lake area.

    4 Results and analyses

    4.1 Comparison of ice thickness estimates

    According to the methods described in Section 3.2, the ice thickness distributions were analyzed and illustrated in Figure 2.According to the ice thickness distribution obtained using GlabTop shown in Figure 2a, the thickness of the Longbasaba Glacier in 2010 was in the range of 0.0 to 288.1±14.6 m,with an aver‐age ice thickness of 48.8±8.1 m. The maximum ice body thickness occurred in the middle of the glacier,and the maximum ice thickness of the ice tongue was approximately 110.1±18.3 m. Figure 2b illustrates the glacier ice thickness distribution acquired using the full-width expansion model. The ice thickness ranged from 0.0 to 172.8±43.9 m, with an average thickness of 55.8±14.2 m. The maximum ice thickness was lo‐cated within the ice tongue, and the thickness was greatest along the glacier's central flowline. The gla‐cier tongue generally extended deep in the middle and was shallow at the upper and lower ends.From the ice thickness distribution derived from the HF-model shown in Figure 2c, one can see that the maximum glacier thickness appeared in the middle of the gla‐cier, with the ice body thickness ranging from 0.0 to 168.5±31.2 m and an average thickness of 29.4±5.4 m.However, the body of the glacier tongue was thicker,and the ice thickness in the glacier accumulation zone was thinner.

    Figure 2 Ice thickness distributions of the Longbasaba Glacier in 2010 obtained using.(a)The GlabTop model,(b)the full-width expansion model,and(c)the HF-model

    The scatter plots presented in Figure 3 illustrate how the estimated ice thickness correlates with the measured. In the legend,nrepresents the number of data points,AD represents the average deviation, and the value in parentheses indicates the percent of the average deviation to the average measured streamline thickness of the glacier. The average deviations of GlabTop, the full-width expansion method, and the HF-model are 9.8, 15.5, and 10.9 m, respectively.The ice thickness uncertainty of GlabTop, the full-width expansion method, and the HF-model is 16.6%,25.4%,and 18.5%,respectively.Accordingly,as Glab‐Top produced the smallest deviation from the mea‐sured value of ice thickness, it was considered better than the other two methods and therefore used to esti‐mate the thickness of the Longbasaba Glacier.

    4.2 Changes in ice thickness and ice volume

    Considering the influence of bedrock wear caused by glacier movement, GlabTop, with its high reliability, was used to estimate the ice thick‐ness of the Longbasaba Glacier in 1980, 2000,2010,and 2015,as shown in Figure 4.The DEM da‐ta used for calculating the ice surface slope in each stage are summarized in Table 1. After the ice thickness of the mid-streamline of the glacier was calculated using GlabTop, the 30 m×30 m resolu‐tion distribution model of ice thickness in each pe‐riod was acquired using Inverse Distance Weighting interpolation.

    Figure 3 Comparison of measured ice thickness of the Longbasaba Glacier in 2010 with thickness estimated using.(a)GlabTop,(b)the full-width expansion model,and(c)the HF-model

    Figure 4 Ice thickness distribution of the Longbasaba Glacier by Glabtop:(a)1980,(b)2000,(c)2010,and(d)2015

    In 1980, an obvious dark closed area existed in the glacier tongue, and a deep depression was visible in the ice bed.The maximum depth was 301.9±50.1 m,and the average ice thickness was 55.1±9.1 m. In 2000, the maximum thickness of the central part of the glacier decreased to 295.2±49.1 m, and the aver‐age glacier thickness fell to 50.7±8.4 m. In 2010, the glacier tongue had thinned further, with the maximum ice thickness at the end of the glacier declined to 122.4±20.3 m, the maximum glacier thickness re‐duced to 288.4±47.9 m, and the average glacier thick‐ness dropped to 48.8±8.1 m. By 2015, the maximum thickness of the glacier accumulation area had de‐creased to 270.5±44.9 m, the maximum thickness of the glacier tongue had diminished to 115.3±19.1 m,and the average glacier thickness further thinned to 47.2±7.8 m. In the 35-year study period, the average glacier thickness declined by 7.9±1.3 m or an average annual rate of 0.23±0.04 m.

    It can be seen that both the area and the volume of the Longbasaba Glacier have decreased consistently over the study period (Table 2). However, although the rate of ice volume reduction has accelerated over time (Figure 5), the area shrinking rate had declined from 2010 to 2015.During this period,glacier thinning was the main reason for the ice volume reduction.

    Table 2 Area and volume of the Longbasaba Glacier from 1980 to 2015

    Figure 5 Changes in ice volume and lake storage

    5 Discussions

    5.1 Sensitivity of model parameters

    Many studies have used these glacier thickness models. For example, Panditet al. (2020) used the GlabTop2_IITB model to report that the uncertainty of 65 glaciers in the Chandra Basin in the Western Hi‐malayas is about ±15.0%. Thickness distributions of the glaciers in the southeastern part of the Qinghai-Ti‐bet Plateau were estimated based on mass balance dis‐tribution data, glacier surface velocity and ice flow principles, and the uncertainties in ice thickness and glacier volume estimates are ±13.2% and 12.8%, re‐spectively (Wuet al., 2020). The glacier volume ob‐tained by the Volume and Topography Automation(VOTA) model ranges between 26.5% underestima‐tion and 16.6% overestimation when compared with the field observations (James and Carrivick, 2016).This study's estimated glacier volume was also com‐pared with a glacier volume dataset derived from a combination of up to five models (Farinottiet al.,2019) and was found to be about 4.0% lower than the average of those models. Rabatelet al. (2012) used surface mass balance and ice flow velocity to estimate the thickness of the French and Argentina glaciers.They found that the ice thickness differs from the ob‐served value by 10% to 30%. Combining the best model and measured ice thickness to obtain the over‐all glacier thickness distribution has also achieved good results (Langhammeret al., 2019). Many re‐searchers compared ground-penetrating radar data with simulated ice thickness (Pieczonkaet al., 2018;Pritchardet al., 2020; Weltyet al., 2020). In this study, it is found that the GlabTop model estimated ice thickness results are in good agreement with the ground-penetrating radar measurements, and the un‐certainty is about 16.6%.

    To investigate the impact of the uncertainty in the model parameters on ice thickness estimation,a sensi‐tivity analysis was performed on the ice thickness esti‐mation results. Each parameter of the different mod‐els (shape factorf, coefficientm, the flow law expo‐nentn, the flow rate factorA, the correction factorC,the surface moraine impact factorfdebris, and the refer‐ence apparent mass balance gradientwas ad‐justed in sequence. In this examination, the estimated resultant change in ice thickness was divided by an undisturbed estimation value of ice thickness to de‐fine "sensitivity" (Δh). For the Longbasaba Glacier,the undisturbed value refers to the mean glacier thick‐ness assessed using the three different models. The outcomes could be applied to predict potential errors in the ice thickness estimations, and these errors would at that point be linked to a certain parameter(e.g., glacier volume) based on the error. Similarly,this parameter would be evident in the different thick‐ness estimation approaches, and its sensitivity investi‐gation was utilized to measure its effect.

    The 2010 data were used as the basis of this sensi‐tivity analysis. The correction parameters in each ice thickness estimation model were adjusted by fitting the ice thickness determined by the model to the GPR-measured thickness.Of the three methods,Glab‐Top and the full-width expansion model have fewer model parameters; the shape factorfand the coeffi‐cientmwere considered. In the validation, two slope parameters (limitα0=4° and thresholdαlim=30°) were adopted (Liet al., 2012). Since there are many model parameters in the HF-model, for simplicity, the most sensitive parameter,i.e., the flow law exponentn, was picked for examination, and the remaining model pa‐rameters were taken from the literature. In a previous study, correction factorCwas set to 0.53 (Farinottiet al.,2009b).The speed coefficient A is generally a pre‐determined estimation of an alpine glacier (Hubbardet al., 1998; Gudmundsson, 1999). The glacier ice densityρwas assumed to be 900 kg/m3,gis 9.8 m/s2,αlimwas set to 20°, andα0was set to 2° (Farinottiet al.,2009a).

    Table 3 lists the model parameters considered in the sensitivity analysis and results.Among the param‐eters considered, it is evident that the ice thickness evaluation is most sensitive tonandf. For example, a 10% change in the values ofnandfwould cause Δhto reduce by 40.7% and 9.1%, respectively. By com‐parison, the ice thickness estimation is much less sen‐sitive to the other parameters, includingm,A,C,fdebris,and db?/dz0.

    Table 3 Sensitivity analysis and parameter setting of ice thickness estimation models

    5.2 Retention of glacier meltwater

    The injection of glacial meltwater into a glacial lake leads to its areal expansion. Furthermore, glacial meltwater does not flow out of the cryosphere entirely through surface runoff but is retained temporarily in the glacial lake. This mechanism helps regulate the water circulation process in mountainous areas to a certain extent and delays regional glacial water re‐sources loss due to global warming (Wanget al.,2013; Das, 2015). Under the background of rising air temperatures and shrinking glaciers in the Himalayas,glacial lakes are anticipated to increase in terms of number and area.

    Based on the topographic map in 1980 and the Landsat images acquired in 2000, 2010, and 2015, it was determined that meltwater from the Longbasaba Glacier had caused Longbasaba Lake to expand mark‐edly from 1980 to 2015. Based on existing data, it is difficult to accurately calculate how much glacial meltwater is retained by glacial lakes. Given the lack of obvious trends in meteorological factors such as precipitation and evaporation, it could be conjectured that expansion of glacial lake is related directly to the injection of glacial meltwater resulting from the gla‐cial retreat and increased melting (Wanget al., 2013).Accordingly, an approximate estimation of the water retained in Longbasaba Lake was provided next.

    Using Equation(10), the net increment of the gla‐cial lake area related to direct injection of glacial melt‐water can be converted into the change of the glacial lake reservoir volume,i.e., the glacial water resource of the Longbasaba Glacier retained by the glacial lake. Based on visual interpretation of topographic maps and Landsat images of different years, the boundaries of Longbasaba Lake in different periods were drawn to assess its area. It was found that Long‐basaba Lake expanded from 0.377 to 1.449 km2from 1980 to 2015. Table 4 summarizes the area and stor‐age volume and its changes in each period.

    Table 4 Changes of area and storage volume of the Lake Longbasaba from 1980 to 2015

    Results in Table 4 show that, from 1980 to 2000,the lake's area increased by 0.51±0.01 km2,its volume grew by 0.0241±0.0082 km3, corresponding to rates of 0.0261±0.0005 km2/a and 0.0013±0.0001 km3/a,re‐spectively, and the relative quantity of retained water in the glacial lake was 16.2%.From 2000 to 2010,the glacial lake area increased by 0.33±0.01 km2, and its storage volume swelled by 0.0112±0.0004 km3, or at rates of 0.0322±0.0006 km2/a and 0.0014±0.0004 km3/a,respectively.The relative quantity of retained water in the glacial lake was approximately 16.7%. From 2010 to 2015,the glacial lake area and volume increased by 0.23±0.01 km2and 0.0109±0.0004 km3, or at 0.0456±0.0009 km2/a and 0.0022±0.0001 km3/a, respectively.And the relative quantity of retained water in the lake was 19.7%. Thus, the changes in the area and storage volume of Longbasaba Lake have accelerated year by year, especially from 2010 to 2015, as illustrated in Figure 5. The increase rates of the lake area and stor‐age volume from 2010 to 2015 were respectively 1.75 and 1.76 times greater than from 1980 to 2000. More‐over, the relative quantity of retained water in the gla‐cial lake has also increased annually.

    5.3 Upward expansion of the proglacial lake

    Change in ice volume and glacial lake storage is often closely related to the regional climatic back‐ground. Some studies have processed and analyzed air temperature and precipitation data obtained at the meteorological stations located near the Longbasaba Glacier (e.g., Dingri Station, Lazi Station, Xigaze Station, Gyangzi Station, and Nielamu Station) (Yaoet al., 2010). It is found that the annual average tem‐perature at the five meteorological stations has shown a warming trend since their establishment, which is consistent with the general warming trend observed on the Qinghai-Tibet Plateau since the 1960s (Renet al., 2004; Yanget al., 2006). Overall, from 1975 to 2015, the annual average temperature in the Kanchen‐junga region, where the Longbasaba Glacier is locat‐ed,increased significantly at a rate of 0.036°C/a(con‐fidence level: <0.001) (Zhaoet al., 2020). In particu‐lar, the temperature increase at the station nearest the Longbasaba Glacier was as fast as 0.37 °C/10a from 1971 to 2009, which is much higher than the global average temperature rise rate of 0.148 °C/10a (Parryet al., 2007). As the warm and humid airflow from the Indian Ocean is affected by the barrier effect of the Himalayas, overall, the precipitation around the Longbasaba Glacier has seen little changes (Wanget al.,2018).

    Some studies concluded that the glaciers on the Qinghai-Tibet Plateau were stable or appeared to even advance slightly during 1970 to 1980, but then started to retreat from the end of 1980 into the 1990s,and the extent and speed of the retreat have increased gradual‐ly in recent years. The Landsat images used in this study show clearly that the boundary of the Longbasa‐ba Glacier has retreated.Over the study period,the re‐treating speed has accelerated, and the area vacated by the retreating glacier had become part of the gla‐cial lake. To summarize, under the background of re‐gional climate warming, the Longbasaba Glacier has recently developed a strong negative mass balance and has been retreating at a rising rate. The lake's re‐tention effect on glacier meltwater is the direct reason for the substantial increase in the area and storage vol‐ume of the Longbasaba Lake.Owing to the lagged re‐sponse of glaciers to climate change, the more obvi‐ous climate warming of recent years will drive the Longbasaba Glacier to continue to retreat, resulting in further expansion of the Longbasaba Lake.

    The basin topography of the Longbasaba Lake/Glacier was constructed, as illustrated in Figure 6.Considering that the glacier front often appears as a steep cliff, a coincident zone was chosen to merge the two overlapping parts of the bottom terrain. It is pre‐dicted that the lake could expand to a distance of 5,000 m from the glacial lake dam(Figure 6).

    Figure 6 The centerline profiles of the glacier bed,ice thickness,and lake depth.V1 represents the lake water volume before the fusion line at the merge location,and V2 represents the lake water volume after the fusion line.V1 is from the glacier bed of 2009,and V2 is from the bed of 1980

    6 Conclusions

    Based on the glacier polygon vector data and the digital elevation model data, three ice thickness distribution models were used to estimate the thick‐ness of the Longbasaba Glacier. By comparison with ground-penetrating radar data, the GlabTop model is found to have the best performance in simulating the ice thickness distribution with an uncertainty of 16.6%. Due to regional climate warming, from 1980 to 2015, the thickness of the Longbasaba Glacier de‐creased by 7.9±1.3 m, with an average annual reduc‐tion of 0.23±0.04 m. During the 35-year study period,the Longbasaba Glacier shrunk by 1.07±0.03 km2,and its ice volume declined to 1.41±0.21 km3as of 2015.

    The growth rate of the area and water storage of Longbasaba Lake has been accelerating. To a certain extent, the expansion of glacial lakes can delay the loss of regional glacier water resources due to climate warming. On average, approximately 17% of glacial meltwater from its host glacier is retained in this pro‐glacial lake every year. The basin morphology of the Longbasaba Glacier/Lake was reconstructed, and the glacial lake may extend from the lake dam by up to 5,000 m.

    Acknowledgments:

    The work was funded by the National Natural Science Foundation of China(No.41701061,No.41761144075,No. 41771075, and No. 41271091). We gratefully ac‐knowledge Zhen Wu and Chao Yao for their help in the ice thickness measurements and processing.We al‐so thank Xuanru Zhao for the ALOS and TanDEM.

    肉色欧美久久久久久久蜜桃| 美女国产高潮福利片在线看| 国产精品一区二区在线观看99| 啦啦啦视频在线资源免费观看| bbb黄色大片| 午夜免费成人在线视频| 午夜免费鲁丝| 亚洲五月婷婷丁香| 国产午夜精品一二区理论片| 老司机亚洲免费影院| 18在线观看网站| 超碰97精品在线观看| 亚洲国产精品一区三区| 免费看十八禁软件| avwww免费| 咕卡用的链子| 成人国产av品久久久| 少妇精品久久久久久久| 国精品久久久久久国模美| 十八禁人妻一区二区| 在线精品无人区一区二区三| 一边摸一边做爽爽视频免费| 一级毛片电影观看| a级毛片在线看网站| 日韩 欧美 亚洲 中文字幕| 亚洲中文日韩欧美视频| 国产精品亚洲av一区麻豆| 久久久精品欧美日韩精品| 国产一级毛片七仙女欲春2 | 一边摸一边做爽爽视频免费| 欧美色欧美亚洲另类二区| 天天添夜夜摸| 久久久久亚洲av毛片大全| 脱女人内裤的视频| 亚洲五月婷婷丁香| 久久狼人影院| 国产精品99久久99久久久不卡| 性欧美人与动物交配| a级毛片a级免费在线| 一级黄色大片毛片| 国内久久婷婷六月综合欲色啪| 男女之事视频高清在线观看| 免费看a级黄色片| 麻豆一二三区av精品| 免费在线观看成人毛片| 国产精品av久久久久免费| 两个人免费观看高清视频| 国产伦在线观看视频一区| 夜夜夜夜夜久久久久| 精品国产乱码久久久久久男人| 国产一区在线观看成人免费| 757午夜福利合集在线观看| 久久久久国产一级毛片高清牌| 男女之事视频高清在线观看| 最好的美女福利视频网| 美女国产高潮福利片在线看| 欧美国产日韩亚洲一区| 麻豆av在线久日| 丁香欧美五月| 亚洲在线自拍视频| 在线观看免费日韩欧美大片| 午夜福利欧美成人| 精品少妇一区二区三区视频日本电影| 午夜老司机福利片| 国产欧美日韩精品亚洲av| 丁香欧美五月| 亚洲一区中文字幕在线| www国产在线视频色| 免费搜索国产男女视频| 他把我摸到了高潮在线观看| ponron亚洲| 黄色a级毛片大全视频| 一进一出抽搐动态| 18禁国产床啪视频网站| 欧美亚洲日本最大视频资源| 人妻丰满熟妇av一区二区三区| 欧美日韩福利视频一区二区| 国产精品av久久久久免费| 婷婷精品国产亚洲av在线| 日本成人三级电影网站| 久久久久久亚洲精品国产蜜桃av| 怎么达到女性高潮| 国产精品二区激情视频| 国产1区2区3区精品| 免费看美女性在线毛片视频| 久久狼人影院| 麻豆成人av在线观看| 一区二区三区国产精品乱码| 精品一区二区三区四区五区乱码| 天天添夜夜摸| 国产精品二区激情视频| 日本五十路高清| 亚洲欧美精品综合一区二区三区| 国产片内射在线| 老司机午夜十八禁免费视频| a级毛片在线看网站| 亚洲精品美女久久久久99蜜臀| 91成人精品电影| 少妇 在线观看| netflix在线观看网站| 夜夜躁狠狠躁天天躁| 欧美黄色淫秽网站| 欧美乱妇无乱码| 欧美黑人巨大hd| 亚洲狠狠婷婷综合久久图片| 亚洲激情在线av| 丁香欧美五月| 亚洲熟女毛片儿| 欧美国产精品va在线观看不卡| 精品久久久久久久久久免费视频| 中出人妻视频一区二区| 99久久国产精品久久久| 一边摸一边抽搐一进一小说| 国产精品自产拍在线观看55亚洲| 香蕉av资源在线| 久久99热这里只有精品18| 久久欧美精品欧美久久欧美| 麻豆成人av在线观看| 好看av亚洲va欧美ⅴa在| 淫妇啪啪啪对白视频| svipshipincom国产片| 精品一区二区三区四区五区乱码| 一本一本综合久久| 久久精品91无色码中文字幕| 一个人免费在线观看的高清视频| 88av欧美| 亚洲片人在线观看| 免费看a级黄色片| 妹子高潮喷水视频| xxx96com| 欧美日韩福利视频一区二区| 久久精品成人免费网站| 国产av一区二区精品久久| 国产午夜福利久久久久久| 国产高清videossex| 欧美一级毛片孕妇| 国产真实乱freesex| 亚洲精品久久国产高清桃花| 性色av乱码一区二区三区2| 1024香蕉在线观看| 一级a爱片免费观看的视频| 国产在线观看jvid| 熟妇人妻久久中文字幕3abv| 夜夜夜夜夜久久久久| 精品国产乱码久久久久久男人| 欧美av亚洲av综合av国产av| 看免费av毛片| 亚洲成国产人片在线观看| 91国产中文字幕| 国产一区二区在线av高清观看| 天天躁夜夜躁狠狠躁躁| 午夜激情av网站| 啦啦啦观看免费观看视频高清| 久99久视频精品免费| 亚洲av电影在线进入| 国产精品亚洲一级av第二区| netflix在线观看网站| 午夜久久久久精精品| 久久久久九九精品影院| 黄片小视频在线播放| 好男人电影高清在线观看| 韩国av一区二区三区四区| 日韩欧美在线二视频| 一个人观看的视频www高清免费观看 | 久久精品国产亚洲av香蕉五月| 亚洲av成人一区二区三| 正在播放国产对白刺激| 嫩草影视91久久| 99在线视频只有这里精品首页| 国产蜜桃级精品一区二区三区| 成人精品一区二区免费| 国产高清videossex| 99久久国产精品久久久| 男女之事视频高清在线观看| 亚洲熟妇中文字幕五十中出| 色在线成人网| 国产99白浆流出| 韩国av一区二区三区四区| 久久久久久久久免费视频了| 色av中文字幕| 一本一本综合久久| 两个人视频免费观看高清| 亚洲av美国av| 亚洲国产毛片av蜜桃av| 一区二区三区高清视频在线| 亚洲一区中文字幕在线| 中文字幕久久专区| 国产久久久一区二区三区| 亚洲人成伊人成综合网2020| 精品免费久久久久久久清纯| 国产不卡一卡二| 一进一出抽搐gif免费好疼| 在线观看www视频免费| netflix在线观看网站| 少妇的丰满在线观看| 久久久久国内视频| 国产精品,欧美在线| 久久国产乱子伦精品免费另类| 亚洲一区二区三区色噜噜| 韩国av一区二区三区四区| 亚洲国产精品久久男人天堂| 日韩高清综合在线| 日韩视频一区二区在线观看| 久久久久久免费高清国产稀缺| 亚洲专区字幕在线| 高潮久久久久久久久久久不卡| 国产亚洲欧美精品永久| x7x7x7水蜜桃| 中国美女看黄片| 一二三四社区在线视频社区8| 亚洲专区中文字幕在线| 久久精品91蜜桃| 国语自产精品视频在线第100页| 国产1区2区3区精品| 色尼玛亚洲综合影院| 最新在线观看一区二区三区| 两人在一起打扑克的视频| 国产区一区二久久| 国产日本99.免费观看| 国产单亲对白刺激| 精品久久久久久久久久久久久 | 精品国产乱子伦一区二区三区| 99精品在免费线老司机午夜| 精品卡一卡二卡四卡免费| 国产免费男女视频| 欧美精品亚洲一区二区| 精品国产超薄肉色丝袜足j| 一区二区三区激情视频| 999精品在线视频| 国产精品久久久久久亚洲av鲁大| 国产av在哪里看| avwww免费| 亚洲av成人av| 成在线人永久免费视频| 国产精品久久久久久人妻精品电影| 日本 欧美在线| 99在线视频只有这里精品首页| 午夜精品久久久久久毛片777| 午夜精品在线福利| 女警被强在线播放| 色综合欧美亚洲国产小说| 欧美中文综合在线视频| 人妻久久中文字幕网| 亚洲欧洲精品一区二区精品久久久| 黄片播放在线免费| 精品高清国产在线一区| 美国免费a级毛片| 变态另类成人亚洲欧美熟女| 欧美亚洲日本最大视频资源| 91字幕亚洲| av中文乱码字幕在线| av欧美777| 一区二区日韩欧美中文字幕| 18禁国产床啪视频网站| 亚洲av日韩精品久久久久久密| 免费看美女性在线毛片视频| 最近最新免费中文字幕在线| 国产亚洲欧美精品永久| 久久国产精品人妻蜜桃| 757午夜福利合集在线观看| 亚洲精品一卡2卡三卡4卡5卡| 国产一区二区三区视频了| 神马国产精品三级电影在线观看 | 岛国在线观看网站| 欧美精品亚洲一区二区| 亚洲成人久久性| 最好的美女福利视频网| 国产主播在线观看一区二区| 国产欧美日韩精品亚洲av| 黄色视频,在线免费观看| 人成视频在线观看免费观看| 在线观看日韩欧美| 国产v大片淫在线免费观看| 最新美女视频免费是黄的| 久久 成人 亚洲| 国产成+人综合+亚洲专区| 国产亚洲精品久久久久久毛片| 神马国产精品三级电影在线观看 | 亚洲欧美精品综合一区二区三区| 亚洲成人免费电影在线观看| 亚洲美女黄片视频| 色综合亚洲欧美另类图片| 亚洲av日韩精品久久久久久密| 久久亚洲真实| 国产成人av教育| 精品久久久久久成人av| 天堂影院成人在线观看| 国产1区2区3区精品| 老熟妇乱子伦视频在线观看| 国产精品亚洲一级av第二区| 韩国精品一区二区三区| 国产精品久久久久久精品电影 | 法律面前人人平等表现在哪些方面| 国产精品电影一区二区三区| 欧美日韩瑟瑟在线播放| 一进一出抽搐gif免费好疼| 久久精品国产亚洲av香蕉五月| 精品久久久久久久毛片微露脸| 18美女黄网站色大片免费观看| 99久久国产精品久久久| 成人18禁在线播放| 久久国产亚洲av麻豆专区| 激情在线观看视频在线高清| 久久久久久免费高清国产稀缺| 村上凉子中文字幕在线| 国产蜜桃级精品一区二区三区| 亚洲性夜色夜夜综合| 一本一本综合久久| √禁漫天堂资源中文www| 国产成人av激情在线播放| 国产熟女xx| 一边摸一边做爽爽视频免费| 人人妻,人人澡人人爽秒播| 久久精品国产清高在天天线| 免费高清在线观看日韩| 高清毛片免费观看视频网站| 很黄的视频免费| 黄色a级毛片大全视频| 国产精品国产高清国产av| 男人舔女人的私密视频| 欧美黑人精品巨大| 嫩草影院精品99| 久久午夜亚洲精品久久| 成人亚洲精品av一区二区| 精品高清国产在线一区| 日本黄色视频三级网站网址| 无遮挡黄片免费观看| 十八禁网站免费在线| 一边摸一边抽搐一进一小说| 亚洲免费av在线视频| 啦啦啦韩国在线观看视频| 亚洲va日本ⅴa欧美va伊人久久| 黄色 视频免费看| 啦啦啦免费观看视频1| 啦啦啦 在线观看视频| 日韩高清综合在线| 成人三级做爰电影| netflix在线观看网站| 成人av一区二区三区在线看| 日韩欧美三级三区| 性色av乱码一区二区三区2| 黄色片一级片一级黄色片| xxx96com| 久久久久国产精品人妻aⅴ院| 波多野结衣高清无吗| videosex国产| 男女视频在线观看网站免费 | 长腿黑丝高跟| 中国美女看黄片| 成人国语在线视频| 精品国产国语对白av| 999久久久国产精品视频| 欧美黄色片欧美黄色片| 1024视频免费在线观看| 99热这里只有精品一区 | 日日夜夜操网爽| 一本大道久久a久久精品| 妹子高潮喷水视频| 在线观看一区二区三区| 香蕉丝袜av| 99精品久久久久人妻精品| 亚洲天堂国产精品一区在线| 亚洲av五月六月丁香网| 日韩成人在线观看一区二区三区| 一二三四社区在线视频社区8| 国产精品av久久久久免费| 首页视频小说图片口味搜索| 在线十欧美十亚洲十日本专区| 伊人久久大香线蕉亚洲五| 他把我摸到了高潮在线观看| 亚洲av熟女| 亚洲欧美精品综合久久99| 国产成+人综合+亚洲专区| 99国产极品粉嫩在线观看| 19禁男女啪啪无遮挡网站| tocl精华| 91老司机精品| 麻豆成人av在线观看| 欧美激情 高清一区二区三区| 这个男人来自地球电影免费观看| 亚洲欧美精品综合久久99| 757午夜福利合集在线观看| 无遮挡黄片免费观看| 午夜福利免费观看在线| 女警被强在线播放| 成人欧美大片| 老司机靠b影院| 一进一出好大好爽视频| 亚洲精华国产精华精| 国产激情久久老熟女| 激情在线观看视频在线高清| 国产不卡一卡二| 色精品久久人妻99蜜桃| 伦理电影免费视频| 国产亚洲欧美98| 亚洲成国产人片在线观看| 国产熟女xx| 日韩欧美国产一区二区入口| 久久久水蜜桃国产精品网| 最近在线观看免费完整版| 亚洲 欧美一区二区三区| 亚洲专区中文字幕在线| 黄色视频不卡| 欧美日韩福利视频一区二区| 成人欧美大片| 伊人久久大香线蕉亚洲五| 色综合站精品国产| 国产欧美日韩一区二区精品| xxx96com| 久久婷婷成人综合色麻豆| www.999成人在线观看| 欧美日韩一级在线毛片| 中文字幕高清在线视频| 欧美+亚洲+日韩+国产| АⅤ资源中文在线天堂| 欧美黑人巨大hd| 色精品久久人妻99蜜桃| 成人午夜高清在线视频 | 免费在线观看视频国产中文字幕亚洲| 淫秽高清视频在线观看| 精品乱码久久久久久99久播| 亚洲七黄色美女视频| 亚洲第一欧美日韩一区二区三区| 成人永久免费在线观看视频| 在线观看一区二区三区| 久久精品成人免费网站| 久久午夜综合久久蜜桃| 成人av一区二区三区在线看| 淫秽高清视频在线观看| 精品无人区乱码1区二区| 免费在线观看视频国产中文字幕亚洲| 激情在线观看视频在线高清| 国产精品久久久久久精品电影 | 在线国产一区二区在线| 日韩精品青青久久久久久| 免费看美女性在线毛片视频| 欧洲精品卡2卡3卡4卡5卡区| 国产精品亚洲一级av第二区| 久久久水蜜桃国产精品网| 久久久精品国产亚洲av高清涩受| 黑人操中国人逼视频| 午夜福利欧美成人| 一级片免费观看大全| 欧美亚洲日本最大视频资源| 日韩 欧美 亚洲 中文字幕| a级毛片在线看网站| 人妻丰满熟妇av一区二区三区| 每晚都被弄得嗷嗷叫到高潮| 久久人妻福利社区极品人妻图片| 两个人看的免费小视频| 久久欧美精品欧美久久欧美| 最近最新中文字幕大全免费视频| 高清在线国产一区| 成人av一区二区三区在线看| 免费观看人在逋| 法律面前人人平等表现在哪些方面| 在线观看www视频免费| 免费高清在线观看日韩| 免费在线观看亚洲国产| 国产精品国产高清国产av| 国产成年人精品一区二区| 老司机午夜十八禁免费视频| 白带黄色成豆腐渣| 亚洲熟妇熟女久久| 国产成人欧美在线观看| 亚洲三区欧美一区| 久久中文字幕人妻熟女| 色尼玛亚洲综合影院| 好男人电影高清在线观看| 亚洲av熟女| 欧美最黄视频在线播放免费| 国产av在哪里看| 高清毛片免费观看视频网站| 91九色精品人成在线观看| 一进一出好大好爽视频| 久久久精品国产亚洲av高清涩受| 美女国产高潮福利片在线看| 亚洲av成人av| 亚洲av电影不卡..在线观看| 精品卡一卡二卡四卡免费| 巨乳人妻的诱惑在线观看| 午夜免费成人在线视频| 嫁个100分男人电影在线观看| 麻豆一二三区av精品| 亚洲美女黄片视频| 国产色视频综合| 久久精品亚洲精品国产色婷小说| 免费电影在线观看免费观看| 国内精品久久久久久久电影| 国产亚洲av高清不卡| 免费看美女性在线毛片视频| 成人18禁高潮啪啪吃奶动态图| 久久人人精品亚洲av| 成人手机av| 两性午夜刺激爽爽歪歪视频在线观看 | www.精华液| 此物有八面人人有两片| av超薄肉色丝袜交足视频| 亚洲av片天天在线观看| 国产精品久久久人人做人人爽| 亚洲成人久久爱视频| 国产精品亚洲美女久久久| 亚洲成av片中文字幕在线观看| 久久草成人影院| 韩国精品一区二区三区| 午夜福利免费观看在线| 亚洲男人的天堂狠狠| 亚洲全国av大片| 丝袜人妻中文字幕| 日韩国内少妇激情av| 久久久水蜜桃国产精品网| 成人三级做爰电影| 91字幕亚洲| 大香蕉久久成人网| 一本大道久久a久久精品| 亚洲狠狠婷婷综合久久图片| 桃色一区二区三区在线观看| 久久久久久免费高清国产稀缺| 久久亚洲真实| 国产伦在线观看视频一区| 国产一区二区在线av高清观看| 高清在线国产一区| 国产精品久久久av美女十八| 国产精品美女特级片免费视频播放器 | 色哟哟哟哟哟哟| 99久久无色码亚洲精品果冻| 精品国产国语对白av| 国产精品二区激情视频| 国产久久久一区二区三区| 2021天堂中文幕一二区在线观 | 国产亚洲av高清不卡| 欧美激情高清一区二区三区| 成人一区二区视频在线观看| 黄色女人牲交| 特大巨黑吊av在线直播 | 久久精品成人免费网站| 国产免费男女视频| 国产av不卡久久| 久久人妻av系列| 国产高清激情床上av| 久久久国产成人精品二区| 男人舔女人的私密视频| 中亚洲国语对白在线视频| 又黄又粗又硬又大视频| 亚洲av日韩精品久久久久久密| svipshipincom国产片| 在线观看www视频免费| 亚洲av日韩精品久久久久久密| 国内精品久久久久久久电影| 欧美国产精品va在线观看不卡| 久久精品夜夜夜夜夜久久蜜豆 | 国产成人av激情在线播放| 欧美一区二区精品小视频在线| 久久久久久国产a免费观看| www国产在线视频色| 亚洲专区国产一区二区| 国产精品一区二区三区四区久久 | 制服诱惑二区| 9191精品国产免费久久| 中文字幕人成人乱码亚洲影| 中亚洲国语对白在线视频| 久久久久久久久免费视频了| 男人舔女人的私密视频| 精品免费久久久久久久清纯| 国产av一区在线观看免费| 每晚都被弄得嗷嗷叫到高潮| 极品教师在线免费播放| 成人18禁在线播放| 午夜视频精品福利| 午夜亚洲福利在线播放| 天堂影院成人在线观看| 午夜福利高清视频| 欧美性猛交╳xxx乱大交人| 熟女少妇亚洲综合色aaa.| 国产成人欧美| а√天堂www在线а√下载| 日日爽夜夜爽网站| 精品国产美女av久久久久小说| 99热这里只有精品一区 | 亚洲成人国产一区在线观看| 一卡2卡三卡四卡精品乱码亚洲| 波多野结衣av一区二区av| 黄色片一级片一级黄色片| 99久久久亚洲精品蜜臀av| 国产亚洲av嫩草精品影院| 侵犯人妻中文字幕一二三四区| 亚洲国产精品成人综合色| 超碰成人久久| 国内毛片毛片毛片毛片毛片| 亚洲中文日韩欧美视频| 19禁男女啪啪无遮挡网站| 成人一区二区视频在线观看| 他把我摸到了高潮在线观看| 法律面前人人平等表现在哪些方面| 91成人精品电影| 99久久精品国产亚洲精品| 波多野结衣高清无吗| 国产成人av教育| 伊人久久大香线蕉亚洲五| 两性夫妻黄色片| 麻豆久久精品国产亚洲av| 禁无遮挡网站| 亚洲电影在线观看av| 一本久久中文字幕| 亚洲国产高清在线一区二区三 | 国产成人欧美在线观看| 久久人人精品亚洲av| 在线av久久热| 亚洲国产精品sss在线观看| 久久久久久亚洲精品国产蜜桃av| 最好的美女福利视频网| 日本撒尿小便嘘嘘汇集6| 精品人妻1区二区| e午夜精品久久久久久久| 久久人人精品亚洲av|