Water storage and drainage of short-lived lakes in the Teskey Range, Central Asia

In the Teskey Range of the Tien Shan (Kyrgyz Republic), four outburst flood disasters from short-lived glacial lakes 10 caused severe damages in the downstream part in 2006, 2008, 2013, and 2014. The short-lived lake grows rapidly and drain within a few months, due to closure and opening of an outlet ice-tunnel in moraine complex at glacier front. The outburst flood of this lake type is a major hazard in this region, it differs from many cases of moraine-dam failure in the eastern Himalaya. To clarify how short-lived lakes store and drain water for short period, we examined its recent changes in water level, area, volume, and discharge with a field survey and satellite data analysis. Korumdu lake appeared and drained within about one 15 month during all summers during 2014–2019 except that in 2016. Water-level data recorded by a data logger and time-lapse camera images show that the lake appeared and expanded suddenly from July to August in 2017–2019. The timing indicates that the lake formed when an outlet ice-tunnel (subsurface channel) drain was blocked by deposition of debris and ice due to ice melting, not by freezing of stored water. Based on calculation of UAV DSMs and water level in 2017, the lake's water volume reached 234,000 m within 29 days, and then the water discharged for 17 days at a maximum rate of 0.66 m/s. The 20 small discharge indicates that the diameter of the outlet ice-tunnel was much smaller than those of four short-lived lakes in the same range that caused large drainages (12-27 m/s) in 2006, 2008, 2013, and 2014. As the results, the dimensions of the outlet ice-tunnel of short-lived glacial lakes presently are related to the flooding scale. Recent warming temperatures may increase both the size of the tunnels and the basin volumes leading to greater hazard from such lakes in the future. In addition, we investigated the timing of appearance of 160 short-lived glacial lakes in this region using Landsat-7/8, Sentinel-2, and 25 PlanetScope satellite images (2013–2018). We conclude that tunnel closure of 117 lakes was due to deposition of debris and ice during summer. The appearance of a short-lived glacial lake is inevitable in summer when the melting rate is high. The characteristics of this lake type might be shown in another Asian mountain permafrost regions. https://doi.org/10.5194/nhess-2020-252 Preprint. Discussion started: 23 October 2020 c © Author(s) 2020. CC BY 4.0 License.


Introduction 30
Compared to the large proglacial lakes in the eastern Himalayas (Ageta et al., 2000;Komori et al., 2004;Bajracharya et al., 2007;Nagai et al., 2017), the northern Tien Shan, Central Asia region instead has many small glacial lakes that are distributed at glacier fronts (Janský et al., 2008;Narama et al., 2010;. Drainage events from these small glacial lakes often produce debris flows and floods. For example, debris flows in 2006, 2008, 2013, 2014, and 2019 in the Teskey Range, northern Tien Shan, caused severe damage including casualties and destroyed bridges, roads, houses, and crops (Narama et al., 2010(Narama et al., , 201835 Daiyrov et al., under review). Such short-lived glacial lakes grow rapidly and drain within a few months (Narama et al., 2010(Narama et al., , 2018Daiyrov et al., 2018). In the Tajik Pamir, drainage from a short-lived lake that formed within 2 years resulted in 25 casualties (Mergili et al., 2012). Such lakes drain through an outlet ice-tunnel (subsurface channel) within an ice-cored moraine complex (Popov, 1987), and are also called nonstationary lakes (Erokhin et al., 2017), though this term also includes lakes with a long lifetime. A short-lived lake can be a severe hazard for local residents because it appears suddenly yet can cause 40 large debris flows. The short-lived lakes are a major hazard in this region, and this differs from the outburst which caused by moraine-dam failure in Himalaya and Andes (Costa and Schuster, 1988;Richardson and Reynolds, 2000;Shreshta 2010;Emmer and Cochachin, 2013;Neupane et al.;.
As such glacial lakes drain through an outlet ice-tunnel, the lake can expand rapidly when the outlet ice-tunnel is blocked due to either freezing or deposition of ice and debris (Narama et al., 2010(Narama et al., , 2018. Drainage then occurs when the outlet ice-tunnel 45 opens during summer. Some short-lived glacial lakes reappear every year (Daiyrov et al., 2018), which is behavior they share with supraglacial lakes on a debris-covered glacier. Several studies have examined the relationship between supraglacial lakes and englacial conduit on a debris-covered glacier (Benn et al., 2000(Benn et al., , 2017Miles et al., 2016;Watson et al., 2016;, but this relationship has seen little study for glacial lakes. Short-lived glacial lakes appear at depressions formed due to glacier recession or the subsidence of either an ice-cored moraine 50 complex (Narama et al., 2010(Narama et al., , 2018Daiyrov et al., 2018) or on a depression formed on a surging glacier (Richardson and Reynolds, 2000;Kääb et al., 2004). Narama et al. (2018) showed that such short-lived glacial lakes tend to appear with three geomorphological characteristics: 1) an ice-cored moraine complex (debris landform containing ice), 2) a depression with a water supply on an ice-cored moraine complex or glacier terminus, and 3) the absence of a visible surface outflow channel from the depression, indicating the existence of an outlet ice-tunnel. 55 Previous studies have argued that the recent expansion of glacial lakes in the Tien Shan is linked to climatic warming and glacier shrinkage (Bolch et al., 2011;Wang et al., 2013;Kapitsa et al., 2017). In addition, Daiyrov et al. (2018) showed that the large variability of glacial lakes was not only related to the local climate condition in the Issyk-Kul Basin, but also to regional geomorphological conditions such as the closure and opening of an outlet ice-tunnel. They also pointed out that icecored moraine complexes have developed under mountain permafrost conditions. Ice degradation within such complex results 60 in moraine formation (Iveronova, 1952;Markov, 1955). Surface changes on an ice-cored moraine complex were confirmed in https://doi.org/10.5194/nhess-2020-252 Preprint. Discussion started: 23 October 2020 c Author(s) 2020. CC BY 4.0 License.
the Jeruy Glacier front between 1979 and 2016 due to ice melting (Daiyrov et al., 2018), and such changes likely affect the outlet ice tunnel and formation of the depressions.
As changes can occur over very large areas and volumes in a short period of time, drainage features and flood scale are extremely unpredictable (Erokhin et al., 2017). Although many short-lived glacial lakes are confirmed in recent years in the 65 northern Tien Shan (Daiyrov et al., 2018), not all short-lived glacial lakes cause large-scale floods. The difference of flood scale remains unclear. A lake's fate depends on whether the dam contains ice (Mergili et al., 2013), and if so, how the outlet ice-tunnel closes and opens. However, the mechanisms of closure and drainage remain also unclear. Such hazards can intensify dramatically and unexpectedly within weeks or even days (Huggel, 2004). In this study, we predict mechanisms of closure and drainage at the Korumdu lake and the reason of different flood scales from shot-lived lakes based on field survey and satellite 70 data analysis. These new knowledges are important for glacier disaster mitigation.
The paper is organized as following. To understand the closure and drainage mechanism of the short-lived lake, we investigated the recent changes in water level, area, volume, and discharge at Korumdu lake based on field survey and satellite data. The Korumdu lake appeared and drained within about one month during all summers in recent years. To clear the reason how the outlet ice-tunnel closes in the Korumdu lake, we examined the surface changes around the Korumdu lake in field survey. To 75 clarify how the other short-lived lakes store and drain water, we investigated the timing of appearance of short-lived lakes for the other lakes in this region were studied in 2015-2019 using Landsat-7/8, Sentinel-2, and PlanetScope satellite images.
Finally, we discussed the reason of outlet ice-tunnel closure at Korumdu lake including other lakes, and the relationship between outlet tunnel size and drainage scale including influence of increasing temperature.

Study area 80
The study area is in the northern part of the Teskey Range and near the south shoreline of the Issyk-Kul Basin, Kyrgyz Republic ( Fig. 1). The glacier distribution in the western part of the range (3700-4200 m) is lower than the distribution in the eastern part (3800-4500 m). The difference is related to differences in annual precipitation, which is higher in the eastern part than in the western part. For example, during 1998-2007, the average annual precipitation at the Kara-Kujur station (2800 m) of the western part is 255 mm, whereas that at the Tien Shan station (3614 m) of the central part is 378 mm, and that at the Chong-85 Ashu station (2788 m) of the eastern part is 550 mm (Podrezov and Ryskal, 2019; Fig. 1). Their annual average temperatures are 0.1°C (1961-1988), -6.28°C (1995-2011Kuzmichenok, 2013), and 0. 27°C (1995-2005), respectively. Recent glacier shrinkage has been smaller in the western than the eastern part of the Teskey Range (Aizen et al., 2006;Narama et al., 2006;Kutuzov and Shahgedanova, 2009).

Field observations at Korumdu lake
The field survey at Korumdu lake was run during the summers of 2015-2019 (Figs. 1, 2). The survey involved measuring the water level and water temperature at the lake bottom with a data logger (Hobo U20) at an interval of 1 h since 21 Aug 2015.
We also set water-level data loggers (Hobo U20) at lake bottom (water pressure) and ground levels (atmospheric pressure).
Water-level logger measurements were corrected to water level (meter) using atmospheric pressure data at the ground. A time-105 lapse camera (Brinto) was also set with an interval of 1 day.
In addition, we obtained aerial images of Korumdu lake-basin acquired by Phantom-4 (DJI) and JABO H601G ( The volume and discharge of the lake in the summers of 2017-2019 were calculated using the daily water level data, UAV 115 DSMs, and time-lapse data on ArcGIS 10.5. The daily volume was calculated based on the 2016 DSM (without water), combined with the DSMs of other years (including amount of glacier recession). We found that the water-level logger measurements agreed with the water levels that were reconstructed from time-lapse camera data based on UAV DSMs. Using the same method, we also reconstructed the water level data between August 4 and 31 based on 10 satellite images from PlanetScope, Landsat-8/Operational Land Imager (OLI), and Sentinel-2 along with UAV DSMs, because we visited at the lake 120 on 4 August 2019. We also investigated the changes in lake area during 2017-2019 using PlanetScope images.
Finally, we examined the climatic and thermal conditions using air and ground temperature data loggers (TR-52i; T&D Co.; resolution accuracy within ± 0.3°C) to log data at 1-hour intervals around the lake. Mean annual air temperature (MAAT) and mean annual ground surface temperature (MAGST) were calculated for 2016-2017.

Timing of appearance of short-lived lakes using satellite data 125
Short-lived glacial lakes were identified using satellite images on ArcGIS 10.5. In particular, 91 images from Landsat-7/Enhanced Thematic Mapper Plus (ETM+, SLC-off) and Landsat-8/OLI, 31 images from Sentinel-2, and 16 images from PlanetScope acquired during 2013-2018. The resolutions of these images are 15 m (pan-sharpened images of Landsat-7/8), 10 m (Sentinel-2), and 3 m (PlanetScope). As a definition of short-lived lake, we use that in Daiyrov et al. (2018), which is based on its seasonal change in area from June to September of each year. Specifically, a short-lived lake is a temporary lake, lasting 130 just one or two years, that appears or doubles in area, then disappears or shrinks within the same year. We counted the number that appeared from June to September each year. In addition, the number was tracked in each given year to examine how it changed from one year to the next. Polygon shapefiles of lakes were extracted manually from the images using ArcGIS 10.3.

Areal variations of Korumdu lake
At the front of the Korumdu glacier lies the Korumdu glacial lake (Fig. 2). It sits in a basin that formed during glacier recession. in 2019 (Fig. 2) due to retreat of the glacier terminus. In the field, we observed ice ridging and debris sliding on the slope of the basin, indicating that the ice was melting from around the shore, thus increasing the width of the basin.
The lake had no discernable surface drainage channel, but we found an outlet point (Fig. 2). The existence of the outlet shows that lake water flows through an outlet ice-tunnel from the lake. The length of the outlet ice-tunnel is 60 m from the entrance of the basin. Drainage water was observed at the outlet point in 2015, 2017, and 2019, but not 2016 and 2018. 145 Concerning size changes, in 2015, the lake appeared in July, becoming large on 30 July, then shrank significantly by 21 August (Fig. 3). For 2017-2019, a more complete story of the changes appears in the images in Fig. 4, which are based on PlanetScope satellite data. The images show that the lake appears suddenly at the end of July to the beginning of August and then shrinks and vanishes by the end of August (Fig. 4a-c). Although the timing of lake expansion differs slightly over the years 2017-2019, the lake always appears in summer. These satellite data demonstrate that the lake is a short-lived glacial lake. 150 The time-lapse on-site images show the same behavior from a different view (Fig. 5). These images also indicate that the lake began to expand from mid-July and reached its maximum level at some time between late July and early August. In contrast, the lake area did not change dynamically in 2016. Based on Landsat-8/OLI data, we also found that the lake appeared in 2014 (May 5, June 27, and September 10). Although these images show rapid drainage, we did not find evidence that the drainage caused flooding during the survey period. According to data in Narama et al. (2018), drainage from Korumdu lake is the flood-155 wave type in the downstream region because the water stream flows on a gentle slope.

Changes in water level, area, volume, and discharge of Korumdu lake
Consider the properties of Korumdu lake from 2017 to 2019. Figure 6 shows the measured water level, lake area and volume, and inflow-outflow discharge. For 2017, we also show the water temperature (Fig. 6a). We also reconstructed the water level 160 data between August 4 and 31 based on 10 satellite images (yellow points in Fig. 6a). We calculated volume and discharge using the water levels and the UAV DSMs.
In 2017, the water level starts increasing from 6 July, reaching a maximum on August 3, and then vanishes on 19 August (Fig.   6a). In the first 29 days, the lake level increases 13 m, the volume reaches 234,000 m 3 , and the area reaches 0.36 km 2 (Fig. 6ac). The resulting rate of volume increase was 8,070 m 3 /day. During discharge, 234,000 m3 of water drains in 17 days, with 165 half of the volume draining from 3 to 7 August 2017 (Fig. 6b) resulting in a maximum outflow discharge of 0.66 m 3 /s (Fig.   6d). Although the water level increases intermittently before August 3, the outflow is relatively smooth. The lake water temperature averaged about 1°C.
In 2018, the water level peaks three times (Fig. 6a). The first, on 25 July, reaches 3.5 m and a volume of 21,000 m 3 (Fig. 6a,b).
The second, the yearly maximum, on 11 August, reaches 6 m and a volume of 53,000 m 3 . Finally, the third peak, on 17 August, 170 reaches a level of 5 m and a volume of 39,000 m 3 . The maximum discharge occurs after the second peak, reaching 0.32 m 3 /s (Fig. 6d). Compared to the case in 2017, the maximum lake level and volume are much smaller in 2018. However, like that in 2017, the inflow rate is also intermittent in 2018. The three peaks indicate that closure of the tunnel occurred several times during the 1-month period.
In 2019, the water level goes up and down until 22 July, when it rises sharply (Fig. 6a). Then the level has a local maximum 175 on 30-31 July, reaching 5 m and a volume of 53,000 m 3 , followed by a yearly maximum on 11 August, reaching 6.5 m and 74,000 m 3 (Fig. 6a,b). The maximum discharge occurs right after the second peak, reaching 0.24 m 3 /s (Fig. 6d).
Considering all three years, the maximum water level is highest in 2017 (Fig. 6a). Over these years, other differences include the timing of the lake-level increases, the number of peaks, and the maximum water volume. All three years had small discharge rates (maxima of 0.66, 0.32, and 0.24 m 3 /s in 2017, 2018, and 2019), consistent with the lack of reported flooding. 180 Concerning fluctuations, according to the water level data for 2017-2019, the level increased with repeated storage-drainage cycles. In the field, we observed sudden small increases of water level in 2016 and 2017, with the level increasing tens of centimeters within 3 h (Fig. 7). These results indicate that water level fluctuations occurred frequently due to closing and opening of the outlet ice-tunnel. and 2018 because the water levels were 3,806.5 and 3,807.5 m, respectively (Fig. 8a,c). These results indicate that the entrance of the outlet ice-tunnel at the basin is at approximately 3,807.5 m, water level too low for drainage. 190 https://doi.org/10.5194/nhess-2020-252 Preprint. Discussion started: 23 October 2020 c Author(s) 2020. CC BY 4.0 License.

Surface changes around Korumdu lake
Over a period of one year, how does the region near the entrance of the outlet ice-tunnel change? To answer this question, we compared UAV orthoimages with DSM data in 2015 and 2016 (Fig. 9). A comparison of Fig. 9a,b shows debris sliding, with horizontal backwasting of an exposed ice ridge by 7 m. The backwasting indicates melting occurred, which is supported by the UAV-derived DSMs in Fig. 9c. In particular, along the profile (a-aʹ; Fig. 9b Korumdu lake is maintained under a permafrost environment. Melting of buried ice causes surface changes including expansion of the lake basin. In addition, melting can lead to closure of an outlet ice-tunnel such as that for a supraglacial lake on a debris-covered glacier Benn et al., 2000;Miles et al., 2016;Watson et al., 2017).

Comparison to other short-lived lakes in the area 205
Korumdu lake appeared during July-August and had relatively little drainage, whereas four other short-lived lakes that appeared in May-June caused large drainages and serious damages (Narama et al., 2010(Narama et al., , 2018. The different appearance times might reflect different processes causing tunnel closure. To help determine how common these appearance times are, we investigated the timing of appearance of short-lived lakes in the northern part of the Teskey Range from June to September during 2013-2018 using Landsat-7/8, Sentinel-2, and PlanetScope satellite images. 210 We identified and examined 160 such short-lived lakes during 2013-2018 (the total includes re-appearances of the same lake in different years) in the study area. In Fig. 10, we classify these by month of appearance. The appearance months with the most lakes are June, the snow-melt period, and July, the ice-melt period; specifically, 43 lakes in June and 90 in July. The total numbers and the proportions of the numbers in these two periods varied during the 6 years. This large variability was not directly related to local climate change (Daiyrov et al., 2018). 215 Concerning re-appearances, 81 lakes appeared only once for 6 years. Of the remaining, 19 appeared twice, 7 appeared 3 times, 2 appeared 4 times, and 2 lakes appeared all 6 years. indicating that tunnel closure occurred with a different month each year.
Short-lived lakes that reappear many years likely have a tunnel condition in which closure occurs easily.

Cause of outlet ice-tunnel closure at Korumdu lake
In the case of ice tunnel closure, the supraglacial lakes on the debris-covered southern Inylchek Glacier in April-May are likely to appear due to the closure of englacial conduits when stored water freezes . The closure of englacial conduits on a debris-covered glacier can be due to roof collapses, creep closure, freezing of stored water, or deposition of ice https://doi.org/10.5194/nhess-2020-252 Preprint. Discussion started: 23 October 2020 c Author(s) 2020. CC BY 4.0 License. and debris (Gully and Benn, 2009;Narama et al., 2017). Collapse of an outlet ice-tunnel wall in a debris-covered glacier can 225 occur by thermal and mechanical erosion Roberts, 2005;Bjornsson, 2010) or by ice deformation (Clague and O'Connor, 2015).
For comparison, in the study region, the outlet ice-tunnel blockages could be caused by the freezing of stored water during winter or by blockage by collapsing with deposition of mixed debris and ice (Popov, 1987;Narama et al., 2010Narama et al., , 2018. In addition, changes in the ice-cored moraine complex due to subsidence or downwasting might cause a blockage in an outlet 230 ice-tunnel (Daiyrov et al., 2018). The short-lived lakes here that caused the four large drainages (2006, 2008, 2013, and 2014) appeared in May-June and expanded in June-July (Narama et al., 2010(Narama et al., , 2018. The timing suggests a closure that is caused by the freezing of stored water during winter or ice-debris deposition (Fig. 11a). We call this the deposition-freezing type.
However, no case was reported in which the tunnel condition and water level fluctuations were compared in detail.
In the case of Korumdu lake, the tunnel closed in July-August of every year since 2014 (excluding the case of no expansion 235 in 2016) based on water-level of a data logger and time-lapse camera images. As we observed changes in the basin on the icecored moraine complex caused by subsidence or downwasting (Fig. 9), the blockages of the outlet ice-tunnel at its entrance or interior likely was caused by deposition of ice and debris due to thermal erosion. This type of blockage is sketched in Fig. 11b.
Further evidence that Korumdu lake forms by the deposition process comes from consideration of water-level fluctuations.
The fluctuations of water level, such as spikes, reveal changes in the tunnel condition (Fig. 6d). A sudden blockage of an outlet 240 ice-tunnel can cause a rapid increase in water level within even a few weeks. For Korumdu lake, the water increase was sporadic, indicating that the outlet ice-tunnel was not completely closed, the blockage was temporary, and the size of the ice tunnel is quite small. As a result, lake drainages can also occur any time in summer, depending on how the outlet ice-tunnel responds to the water pressure or thermal erosion.
In 2017, the trend of water volume increase consisted of two parts: 5 to 25 July and 26 July to 3 August (Fig. 6b). The first 245 period had sporadic fluctuations, indicating incomplete closure of the tunnel, but the second period had a smooth increase, indicating complete closure. The lake area reached its maximum value in 2017. This indicates that the period of tunnel closure was longer in 2017 than in 2018 or 2019. Longer closure periods are associated with the formation of larger short-lived lakes (Narama et al., 2018). Thus, the period of closure might be determined by the condition of the tunnel.
Many of the other short-lived lakes that also appear in the ice-melting period are likely to be the deposition-closure type, for 250 the same reasons we applied to Korumdu lake. For example, in Fig. 12, we show surface changes in the outlet ice-tunnel at the Jeruy glacial lake between 2014 and 2016. Ice melting caused large changes and rapid deposition within the outlet ice-tunnel, making closure likely. Thus, the surface condition always changes on an ice-cored moraine complex within the mountain permafrost zone, and the deposition-closure type is the major type in this region. Thus, the appearance of a short-lived glacial lake is inevitable in summer when the melting rate is high. The characteristics of this lake disaster might be shown in another 255 Asian mountain permafrost regions.
Of the 160 short-lived lakes we identified in 2013-2019, only Jeruy lake (in 2013) and Karateke lake (in 2014) had large drainages. The estimated maximum discharges from Jeruy (182,000 m 3 ) and Karateke (123,000 m 3 ) lakes were 14.9 and 11.5 260 m 3 /s, respectively (Narama et al., 2018). These lakes had relatively large outlet tunnels, with Jeruy's cross-section being 4 x 2 m 2 (Fig. 12a,b) and Karateke's about the same or larger (not shown). Earlier, back in 2008, the w-Zyndan lake of 437,000 m 3 had a discharge rate of 27 m 3 /s (Narama et al., 2010). Most of the water in these three cases drained over a period of several hours. In contrast, Korumdu lake did not have a large drainage in 2014-2019, and its tunnel cross-section was much smaller than those of Jeruy or Karateke lakes. For example, in 2017, 234,000 m 3 drained from the lake over 17 days, with a maximum 265 discharge rate of 0.66 m 3 /s, about 20 times smaller than that of the two large drainages of Jeruy and Karateke.
In addition, with Korumdu lake we observed sudden fluctuations of water level over several hours, which is behavior consistent with closure of a small channel caused by deposition. The relatively small tunnel size of this lake ensured a slower discharge even when it became full (330,000 m 3 ). During 2017-2019, the lake size was largest in 2017, yet the discharge rates were nearly the same every year. These results show that the lake size and the dimensions of the outlet ice-tunnel are related to the 270 scale of discharge.
However, tunnel dimensions could increase in the future due to thermal erosion, allowing greater discharge rates. Meltwater and increasing temperature can accelerate thermokarst processes Kääb et al., 2001;Miles et al., 2018), enlarging the outlet ice-tunnel. In addition, although basin-size changes depend on the particular glacier landforms, the basin area in the case of Korumdu lake has increased each year due to glacier recession. If this applies to other short-lived glacier 275 lakes in this region, large-scale flooding events during their discharge may increase in the future due to increasing temperature.

Conclusion
Our field survey found that Korumdu lake appeared and expanded from July to August and then drained over a period of 2-3 weeks. The lake formed when its outlet ice-tunnel closed, which was due to deposition of debris and ice during summer. Later, 280 the draining process was relatively slow because the outlet ice tunnel was small and the scale of discharge is related to the sizes of the outlet ice-tunnel and the lake volume. We argued that predicting the scale of a drainage requires knowledge of the outlet ice-tunnel dimensions and the lake's depression size. Our research method of combination between water-level data and UAV DSMs could estimate the discharge and the approximate dimensions of the tunnel.
During 2013-2018, satellite data showed this region to have 160 short-lived glacial lakes, many of which had a similar timing 285 of appearance as Korumdu lake. Four lakes that appeared a month earlier had large drainages, the only cases of large drainage in the study. Nevertheless, with a warming climate, any short-lived glacial lake might cause large flooding if the outlet ice tunnel and basin size sufficiently enlarge.
The glacial lake outburst floods (GLOFs) which caused by moraine-dam failure such as Himalaya and Andes are minor cases in this region. Short-lived lakes which caused by closure and opening of an outlet ice-tunnel in moraine complex are a major 290 hazard in this region, because the short-lived lake exists on an ice-cored moraine complex within geomorphological and climate conditions of the mountain permafrost zone. In general, short-lived lakes should be monitored using satellite data and field https://doi.org/10.5194/nhess-2020-252 Preprint. Discussion started: 23 October 2020 c Author(s) 2020. CC BY 4.0 License.
observations. These new knowledges are useful to understand the phenomena and behavior of the short-lived lakes and consider glacier hazard mitigation in the mountain permafrost regions of Asian high mountains. A threat of the short-lived lakes increases for the residents since 2000s. This hazard case might be major in Asian high mountains in present. 295    ice-cored moraine complex in the case of a short-lived lake (based on Popov, 1987). (a) Deposition-freezing type of closure that appears when an outlet ice-tunnel is blocked due to the freezing of storage water or deposition of debris and ice. (b) Deposition-closure type that appears when an outlet ice-tunnel at the entrance or interior is blocked due to deposition of debris and ice by thermal erosion (ice melting).