The Sikkim flood of October 2023: Drivers, causes and impacts of a multihazard cascade

PART-III

Future GLOF hazard in the Teesta Valley

SLL remains highly susceptible to future GLOF events, including repeat triggers from northern lateral moraine failures. Despite the 3 October failure and associated slope changes, the northern moraine still comprises a large and rapidly deforming zone. We computed post-event surface velocities using 1635 satellite image pairs between October 2023 and June 2024, revealing that a ~0.5 by 0.3 km region of the collapse scarp is deforming at rates up to 15 m a^?1 (fig. S44). The modified slope geometry following the collapse may cause further failures, with moraine curvature at the crest now higher than before the 2023 failure. Small-scale mass movements are visible on the failure slope (fig. S45 and methods section “Pre- and post-GLOF dynamics of the lateral moraine”).

Debuttressing due to glacier surface lowering and glacier retreat must be considered a primary factor for slope destabilization, increasing outward and downward forces in the frozen moraine. SLL is expected to grow by another ~1 km in length as the glacier retreats (14, 49). With continued retreat of the calving front, debuttressing will affect frozen moraine slopes up-glacier (zone 1), which already show slow downslope movement post failure (fig. S44B and methods section “Pre- and post-GLOF dynamics of the lateral moraine”). Lateral stress coupling must have induced load removal on the up-glacier part, causing it to slow down. A GLOF could potentially be triggered by exposure of Zone 1 on the northern lateral moraine, particularly the eastern flank, due to loss of lateral support following the 3 October collapse. As well as this slope debuttressing, steep slopes surrounding the lake are potential avalanche source zones and thus potential GLOF triggers at moraine-dammed lakes (14, 50). The southern moraine appears stable. However, continued warming, glacier retreat, and permafrost decay could initiate instability in the northern moraine. Downwasting of exposed dead ice on the breached frontal moraine could lower the lake's outlet channel, increasing outflow during future GLOF events.

The GLOF eroded the riverbanks laterally, weakening them and making them susceptible to future collapse, particularly near roads and settlements. For instance, post-GLOF landslide (L17) and slumping below Lachen (see fig. S35) show widened riverbank scarps encroaching closer to settlements. The Naga landslide (L43) also showed slumping in the months after the GLOF (fig. S34). Significant lateral erosion damaged the national highway (NH-10) in multiple locations (fig. S32), blocking major trade routes and isolating mountain communities. The ongoing deterioration of roads months after the 3 October GLOF event, exacerbated by subsequent monsoon floods, further eroded the valley walls, posing a hazard to infrastructure and disruption to transport (fig. S47).

(Fig. 6. Field evidence of sediment aggradation.(A to F) Photographs taken along the Teesta River show the aggradation of the sediments transported by the flood cascade and its impact. Latitude, longitude, and elevation (in m a.s.l) are at top right; locality name and distance from SLL are at bottom right. Photo credits: Praful Rao (co-author).

Flood deposits along the Teesta Valley remain exposed to further erosion and transport, potentially triggering future debris flows (Figs. 5 and 6). Moreover, aggradation has raised the riverbed by several meters, heightening the risk of early onset of bank-full conditions during future floods, increasing the probability of flooding in adjacent floodplains, and exposing populations and infrastructure to greater risks (fig. S46). This concern extends to future GLOFs and high discharge, monsoonal flood events. Crucially, even though the landslide-dammed lake (L6) formed after the GLOF event partially drained, the landslide deposits still present a continuing hazard, potentially amplifying the impact of future GLOFsoriginating upstream (figs. S28M and S29). These eroded sediments are rarely considered in the analysis of GLOF risks.

Summary and perspectives

The multihazard cascade and consequent disaster of 3 October 2023 underscore challenges in GLOF and multihazard assessments that often underestimate the potential intensity and impacts in mountain regions where the hazard from the GLOF itself is significantly conditioned, and in this case, exacerbated, by the downstream geomorphic system (51). The SLL triggering was not remarkable in terms of rainfall; rather, the situation was significantly exacerbated by the effects of climate warming on the drivers of GLOF. On 3 and 4 October, the Teesta Valley experienced heavy rainfall, which saturated the soil and increased the vulnerability of slopes to failure. This preconditioning effect primed the landscape, leading to numerous landslides triggered by the GLOF event. These secondary landslides added to the sediment volume in the floodwaters and contributed to the overall devastation along the downstream flow paths. Rainfall fueled the flood cascade downstream. This additional influx of water intensified the volume and velocity of the floodwaters, leading to more severe impacts on infrastructure, communities, and agricultural lands in Sikkim, West Bengal, and Bangladesh.

The sheer volume of water (~50 × 10⁶ m³) released from the lake, together with the sediment (~270 × 10⁶ m³) entrained along the valley drove the primary impacts that overwhelmed infrastructure and developmental activities along the Teesta River, exacerbating the human and economic toll. Despite the Teesta-III hydropower reservoir contributing 5 × 10⁶ m³ of water (assuming it was at full capacity), which is 10% relative to the initial SLL outburst volume, the GLOF's volume and especially its eroded sediment load dominated downstream impacts. Prevailing GLOFmodeling and assessment approaches insufficiently account for processes of erosion and sediment transport, as well as hillslope-channel interactions such as riverbank collapses and landslides triggered by toe-undercutting as well as the impact of sediment transport on local bed elevations and hence water levels. The latter is of particular importance in large river basins because water waves move faster than sediment waves (52), with eventual deposition therefore driven by not only changing exogenic forcing (e.g., reductions in valley slope) but also endogenic processes where water outruns sediment. These processes alter flow rheology along GLOF tracks and thus flow behavior and geomorphic impact (53, 54), yet adequate tools are lacking to support modeling, simulation, and prediction. Based on our calculation from DoD and GLOF volume, the ratio of the mobilized sediment to the water released from SLL and the Chungthang reservoir reaches 0.83 at the downstream end of the erosion zone. The calculated lake outburst volume and sediment entrainment along the flow path indicate a bulking factor of about 5 (i.e., a 5 times increase in flow volume) which is at the upper end of comparable large debris-laden flows (such as GLOFs, debris flows, lahars) (55, 56). Erosion rates averaged over 70 km to SLL are ~3850 m³ m^?1 (Fig. 5) which is three orders of magnitude higher than observed for granular alpine debris flows. There, intense precipitation the days prior to the GLOF has likely played an important role in the very high erosion and entrainment processes by wettening and saturating the soil along the flow path, as flow conditions and bed wetness are decisive factors to control erosion (57). Neglecting intense sediment entrainment and subsequent bulking (and dilution) can lead to inaccuracies in flood models, potentially underestimating the hazard posed by GLOFs and meaning that design standards for infrastructure may not be appropriate. Hence, comprehensive and integrative approaches to GLOF hazard assessment (3) are urgently needed, considering not only the lake and outburst potential but also downstream landslide susceptibility along the flow path and potential for cascading processes. Also evaluating geomorphic work induced by these GLOF events relative to normal monsoonal floods has scope for future assessments.

This Sikkim flood event is a reminder of some much wider implications including the urgent need for Early Warning Systems (EWS) in the Himalaya, recognizing the complex technical, practical, institutional, and social dimensions that need to be addressed. Expanding and enhancing these systems across the Himalaya is critical for timely hazard detection and effective response, as well as reducing the impact of future GLOFs on communities and infrastructure [c.f. (58)]. Addressing these complexities requires robust infrastructure, advanced technology, and effective coordination among stakeholders (59) to ensure the reliability and effectiveness of EWS in the Himalaya and other challenging mountain environments. In terms of transboundaryGLOF impact, this event demonstrates the complex and interconnected nature of natural hazards in mountainous regions and their far-reaching damage, highlighting the importance of regional cooperation and coordinated efforts among countries sharing river basins to enhance resilience and preparedness against the increasing risks posed by GLOFs (26, 58, 60). Moreover, the significant impact of intense precipitation on flood dynamics and downstream effects observed during this event, particularly in Bangladesh, highlighted the urgent need to integrate response planning and enhance preparedness from a transboundary perspective.

Efforts to mitigate the hazard posed by SLL have been ongoing before the catastrophic flood. An initial lake bathymetric survey was conducted in August 2014, and the first mitigation measures began in September 2016 through the installation of siphons to lower the lake level (61). The most recent expedition was in September 2023, just before the lake's outburst on 3 October, when repeat bathymetric measurements were conducted, and an automated weather station and cameras were installed at the lake site (62). The expedition also recommended additional mitigation measures, such as constructing check dams, retention walls, deflection dams, and implementing anEWS (34) in the valley. In light of the consistently high hazard levels in SLL and valley conditions following the October 3 GLOF event, which has caused rapid remobilization of flood sediments, urgent risk mitigation and management plans are required. These plans must address the altered conditions of both the lake and valley and prepare for potential future scenarios. Comparable conditions were noted right after the Chamoli event (63). While the 3 October disaster has placed the immediate focus on SLL, broader attention, and high priority also needs to be given to the various potentially dangerous lakes identified across High Mountain Asia region. The need for enhanced basin-scale EWS, adaptive infrastructure planning, and cross-border collaboration in hazard management is evident to mitigate the socio-economic and environmental consequences of future GLOF events.

Strengthening regulatory frameworks is crucial to mitigate the increasing risks posed by the proximity of hydropower projects to glacier lakes and in high mountain environments in general. The trend of high GLOF susceptibility in the Himalaya indicates a greater likelihood of future GLOFs, exacerbated by the growing number of hydropower projects moving closer to these hazard-prone areas, thereby increasing exposure. With 47 hydropower projects and an installed capacity of >5300 MW, the Teesta basin has the highest density of such projects in the Himalayan region (64). These numbers are likely to increase and thus, comprehensive risk assessments, stringent building standards, and adaptive management practices are essential to ensure safety and sustainability in these vulnerable regions. This is crucial for safeguarding both infrastructure investments and the communities reliant on these developments in the Himalaya and other mountain ecosystems. Events of the magnitude of the South LhonakGLOF, Chamoli ice-rock avalanche of 2021 (27), or Kedarnath flooding of 2013 (23) highlight potential limits to adaptation in the Himalaya, with even the most diligent and comprehensive suite of disaster risk reduction strategies unlikely to entirely prevent losses and damages occurring from such events. This calls for adequate assessment and communication of residual risks, and effective risk transfer mechanisms, such as insurance and governmental support, to ensure sustainable mountain development. This study highlights the necessity to establish specific guidelines and standards for GLOF risk reduction in the Himalaya and similar high-mountain regions. Structural and non-structural GLOF mitigation strategies should be prioritized, using advanced technology to address risks in extreme climate regimes.

The 3 October 2023 GLOF from SLL highlights the urgency of a paradigm shift in numerical modeling and observational techniques for GLOFs. This urgency extends to improving GLOF risk management and infrastructure development in high mountain regions. These shifts in approaches should help safeguard against the devastating impacts of GLOFs, thereby facilitating sustainable development in hazard-prone environments globally. We contend that improved EWS coupled with enhanced infrastructure resilience and rigorous land-use management practices are essential to mitigate GLOF risks. Furthermore, robust community preparedness and education programs are crucial for effective emergency responses. This multihazard cascade exhibits the complex interactions between climate change, glacier mass loss, and human infrastructure in mountainous regions. Understanding and addressing multihazard cascades in similar vulnerable environments requires interdisciplinary approaches, robust monitoring systems, and proactive measures to minimize devastating consequences and enhance resilience.

TO BE CONTINUED

(https://www.science.org/doi/10.1126/science.ads2659?fbclid=IwZXh0bgNhZW0CMTEAAR2Xx8WLq9ZqVg_UhG3M2O2UZCFz5kBGOtSJ0TLxsZhNeYJytkyNaH7bH8M_aem_XDjyoEsPLbbclUL2C8-G3w)