Research Dossier:
Mapping the Environmental Impacts of the Siachen Glacier Military Occupation
Hamza Jilani
March 2024
Data/Evidence + Technical Resources
GLIMS Glacier Database
GLIMS stands for the Global Land Ice Measurements from Space and it is a data collection from an international project to inventory the world's glaciers and enable scientists to map how glaciers have changed over time, “allowing them to better understand the impacts these changes will have on water resources, downstream hazards, ecosystem changes, and global sea level rise.” Their publicized applications are: global change detection, hazards detection and assessment, and glacier monitoring; ultimately, there is an explicit ecological angle to GLIMS. Correspondingly, the glacier database includes measurements of glacier geometry, glacier area, snowlines, supraglacial lakes and rock debris, and other glacial attributes, as well as browse images. Thus, this data is relevant for understanding and modelling the ecological impacts of glacier extent.
70 percent of the world's glaciers are covered by this data, which is primarily derived from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) instrument aboard the Terra satellite and the Landsat Enhanced Thematic Mapper Plus (ETM+). A joint U.S./Japan ASTER Science Team established the GLIMS initiative in 1999.
Digital glacier outlines, related metadata, and literature references, as well as – more recently – snow lines, center flow lines, and hypsometry data are used for analyzing and optimizing glacier imaging. The visual assemblage of satellite remote sensing has been traced back to the early days of the Cold War, and it co-evolved with knowledge controversies on environmental security and a mainstream planetary gaze that has been a part of efforts to develop a comprehensive global monitoring system. According to Matthew Farish in The Contours of America’s Cold War, social sciences were essentially militarized in the early Cold War period, producing geospatial knowledge that was a form of power and especially of use to the the U.S. for its military deployments, diplomacy, espionage, and finance. Given that the Cold War activities of the United States are tied up in the geopolitical context of Siachen, especially through American and Pakistani involvement in Afghanistan, these Cold War politics evoke the direct lineage of the territorial issue itself as well.
Finally, despite the explicit ecological angle to GLIMS and its emphasis on risk and vulnerability, the overall neoliberalization of remote sensing has happened alongside the shifting of environmental security towards human security and resilience. This becomes problematic when it manifests itself together with the dismissal of attempts to prevent hazardous environmental change in the first place and the inhibition of efforts to carry out non-market-based, community-oriented risk management that might not be as governable or analyzable by the regime of visual assemblage of satellite remote sensing. Moreover, this neoliberalization contributes towards capital-dependent inequities in communities’ abilities to afford environmental intelligence products that might be able help them adapt to climate change-related disasters (Delf Rothe, 2017).
Randolph Subset of GLIMS Glacier Database
The Randolph Glacier Inventory (RGI) is a globally complete inventory of glacier outlines and a subset of the GLIMS database. Thus, my already-outlined reflections for GLIMS apply. However, unlike GLIMS, which is multi-temporal and has many attributes that can be used to model ecological change over time, the RGI is a snapshot of the world's glaciers at a specific target date. Despite being updated many times out of a prioritization of achieving global coverage, consistency, and proximity in a specific year, RGI’s polygon dataset represents a snapshot of what glaciers looked like in 2000, suggesting that global comparisons between glacial shapes as they are mapped at a specific point in time as well as the estimation of volumes and rates of elevation change at regional and global scales and the simulation of cryospheric responses to climate shocks appear to be central foci of this data set. Thus, this subset allows us to think about large-scale global changes and differences between glacial areas, but local differences are less pronounced.
This is an academic contribution to the cultural “gigantification” of glaciers as things to be analyzed on large scales or as global or regional resources, rather than actors in more localized issues, as my project attempts to do. I still do, however, want to recognize glaciers’ relationship to literal downstream ecological effects such as polluted runoff that could have the kinds of regional or global impacts that RGI prioritizes: for instance, they claim that RGI data can simulate cryospheric responses to climatic forcing. Moreover, RGI’s static nature may also be considered a manifestation of the emphasis on “gigantification” that has endured within the discipline of cartography and influenced the satellite data collection techniques that enable it today – mapping has, throughout its history, been a form of political and economic standardization that serves to make territory legible and manageable by authority (Delf Rothe, 2017).
Thus, despite using RGI to build out my base layer that shows the glacier as a whole, I aim to include local context as well as a wider breadth of imagery to combat the oversimplification of Siachen into simply the polygon that its glacial outline creates.
Sentinel-2 L2A
Sentinel 2 is a satellite project that has the goal of "providing systematic global acquisitions of high-resolution, multispectral images allied to a high revisit frequency" and attempts to contribute to climate change, land monitoring, emergency management, and security research. Thus, its data, too, is tied up in Cold War politics and the neoliberalization of remote sensing and environmental risk management.
SENTINEL-2 has 13 spectral bands and a 290 km swath width as well as a high revisit frequency; thus, it uniquely provides valuable data for land cover/change classification, atmospheric correction, and cloud and snow masks, presumably due to its declared focus on measuring the effects of climate change.
Digital elevation models (DEMs) derived from historical topographic maps and aerial photography can be used as baseline for comparison with more recent satellite-derived DEMs, but in the Indian Himalayas, this can prove tricky. Thus, in addition to governmental barriers including restrictions on the export of maps, aerial photographs, and trigonometric and gravity data, “restrictions on aerial photographs and topographic maps at scales larger than 1: 100,000 from areas 80 km wide along the external land border and coastline” cause DEMs of the region to be sparse. This can seriously limit what can be discerned from the Sentinel-2 data in terms of modeling global climate and environmental change. Such restrictions do not exist in the Nepal Himalayas, where large scale topography is allowed and DEMS can be used for glaciological studies (Racoviteanu, Williams, and Barry, 2008).
Landsat/Airbus/Maxar (via Google Earth)
The broader view timeline in my project uses imagery from Landsat gathered via Google Earth; additionally, in order to geographically locate points of interest, I used imagery from Airbus and Maxar gathered via Google Earth, since it actually provided more detailed imagery, especially of urban forms and military base infrastructure, than Sentinel and GLIMS.
This detail is possible because the map imagery in Google Earth is a cloudless composite made from satellite data. Thus, as a quilt of many preprocessed images stitched together, the representations in Google Earth are not ‘real’ images, tying this dataset up in trajectories of data ‘gigantification,’ where certain aspects of images are privileged over others in service of a less holistic composite that can be used to make broad observations with less attention to certain kinds of information, some of which may be lost through the creation of the composite image. The resulting timelapses made with this imagery are thus not fully representations of change over time but instead representations of changes in these ‘average’ representations over time.
Moreover, Landsat, Maxar, and Airbus are all datasets in the service of defense, intelligence, and private industries. Landsat’s development was in service of the U.S.’s efforts to catalog Soviet territories and their capacities for crop production in order to assess the support that would be given to various anticolonial independence movements, underscoring this satellite data’s direct involvement in the U.S.’s imperial project during the Cold War. Similarly, Airbus currently aims to “increase security, optimize mission planning and operations, improve management of resources, [and] boost operational performance” with its data while Maxar lists government, including military, and commercial partners as the primary parties it serves. These foci lead to publicly available high-resolution true color imagery of Siachen glacier, but these capabilities are linked to their applications for resource management and military intelligence. Thus, in using these datasets, I am interacting with a visual assemblage whose capabilities have been shaped by their service of imperialism and capitalism. However, the publicization of data from these imperial-linked satellites allows me to access data that might be similar in nature to the classified Indian military data regarding Siachen.
Timemap (Technical Resource)
Timemap is a software developed by Forensic Architecture to dynamically visualize time and space events from a spreadsheet database in a browser by proxying source data in a spreadsheet and making it available as JSON endpoints. In order to build a project out as a Timemap, I would have to clone and install timemap, configure it using the provided examples in the Forensic Architecture guide and run it locally, then modify a corresponding data sheet to include my events of military-environment interactions at Siachen Glacier. Timemap is used by FA to present research findings online and in exhibitions but also throughout all stages of an investigation, as the density or scarcity of events in the axes of time and space in a Timemap informs where FA “ought to next focus their research, where and when data resolution is high, and where and when we need more moments.” Thus, in addition to providing important time-based context to geospatial representations, Timemap can improve the quality of depth and focus within my investigation as it is in progress by providing insights into what should be investigated further.