Climate Change in Gilgit-Baltistan: Why Indigenous Knowledge Still Outpaces Costly Warning Systems
By Salma Khan
The fragile ecology of Gilgit-Baltistan is at the forefront of climate change. Glacial lake outburst floods (GLOFs), driven by melting ice and rising temperatures, have devastated communities in recent years. The rapidly increasing temperatures in high-altitude mountain regions, such as the Hindukush, Karakoram, and Himalaya (HKH), are causing extreme catastrophes. There are approximately 54,000 glaciers in the HKH, covering an area of 60,000 sq km. Among these, the Himalayas host 60-70% of glaciers, the Karakoram 28-33%, and the Hindukush 10-15% due to its smaller glaciated area (Bajracharya et al., 2011; Nuimura et al., 2015; Wester et al., 2019).
Hunza and Ghizer Valley are the epicenter of GLOF in GB. Ghizer Valley, the northernmost region, shares a border with Afghanistan and China. Climate projections indicate a decline in western disturbances combined with rising temperatures, which may increase the frequency and intensity of extreme precipitation events and trigger large-scale debris flows in the region (Immerzeel et al., 2014). According to a recent study (Mazhar et al., 2025), in the last 200 years, 24 GLOF events have occurred in the Ghizer district alone. Under the UNDP GLOF-II, Early Warning Systems (EWS) have been installed in 16 highly risky valleys in GB. These EWS include automatic weather stations, rain gauges, and water level sensors, which are designed to provide timely alerts to the local communities before GLOF and climate-induced disasters. The installation of EWS in GB is a collaborative project involving both governmental and international organizations, but the primary body responsible is Pakistan’s Ministry of Climate Change and Pakistan Meteorological Department (PMD).
On August 22, 2025, a glacier burst in Roushan Valley, Ghizer, formed a debris-dammed lake that later breached, resulting in a landslide-dammed lake outburst flood (LDOF), often misclassified as a GLOF. The LDOF caused widespread destruction in the region, destroying everything in its path. Interesting, Roushan Valley is not included in the UNDP and the government GLOF projects under which EWS were installed. The absence of EWS highlights a critical gap in hazard preparedness. The government authorities were seen taking credit for the zero casualties. However, these were solely due to community-based warning. A local shepherd spotted the danger at 3 a.m. and alerted the locals just a few minutes before the flood. While Hundreds of lives were saved, more than 300 houses were destroyed.
This unprecedented event in Roushan Valley raises several critical questions, such as why no EWS was installed in a historically high-risk GLOF zone? Why do expensive government-installed early warning systems so often fail, while indigenous knowledge continues to save lives? Despite repeated GLOF events in Ghizer and Hunza, government-funded EWS often fail when most needed. Studies indicate that many systems in northern Pakistan are technologically insufficient, poorly maintained, or unable to operate under extreme weather conditions (Din et al., 2023; Ahmed et al., 2021). Additionally, the regions’ topography, frequent power outages, weak data transmission infrastructure, and a lack of trained local operators further compromise their effectiveness. Consequently, high-cost installations often become symbolic rather than functional.
The glaciers of the Karakoram are known for their anomalous behavior, advancing, retreating, or surging unpredictably compared to neighbouring glaciers in the Hindukush and Himalayas. Local communities, through generations of close observation, are deeply aware of these patterns and can often anticipate glacier-related hazards far more effectively than mechanized systems. Indigenous communities have long relied on experiential knowledge (observing glacier sounds, unusual river flows, livestock behavior, and local weather patterns) to predict floods (Ali & Shah, 2022). This knowledge is continuously updated, deeply contextual, and remarkably adaptive to changing environmental conditions. Yet, official disaster management programs rarely incorporate these insights, despite their potential to improve timely and context-specific early warning. With over 3,000 glacial lakes in Gilgit-Baltistan, 33 classified as high-risk, there is an urgent need to integrate community expertise into EWS planning (Shrestha et al., 2022).
The Ghizer tragedy demonstrates that technology alone cannot ensure resilience; hybrid systems combining modern sensors with local knowledge could provide more reliable alerts. Policymakers must address the gaps highlighted in hazard preparedness. Without these reforms, reliance on expensive, underperforming machines will continue to erode public trust, leaving communities dependent on traditional observational methods that have safeguarded lives for generations.
Salma Khan is a researcher and PhD scholar whose work focuses on the intersections of modern science and indigenous knowledge of climate change in Gilgit-Baltistan. Originally from Aliabad and currently based in Milan, Salma has conducted extensive fieldwork across Gilgit-Baltistan, documenting traditional ecological practices such as glacier grafting and the role of shamans in sustaining local environments.
