Using Drones to Study Glaciers

Understanding the nature of glacial changes has become increasingly important as anthropogenic climate change alters their pace and extent. A new study published in The Cryosphere Discussions journal shows how Unmanned Aerial Vehicles (UAVs), commonly known as drones, can be used to do this in a relatively cheap, safe and accurate way. The study represents the first time a drone has been used to study a high-altitude tropical Andean glacier, offering insight into melt rates and glacial lake outburst flood (GLOF) hazards in Peru.

The researchers used a custom-built drone (Source: Oliver Wigmore).

The study was carried out by Oliver Wigmore and Bryan Mark, from the University of Colorado Boulder and Ohio State University respectively. It is part of a larger project aimed at understanding how climate change is affecting the hydrology of the region and how locals are adapting to these changes.

The researchers used a custom-built hexa-multirotor drone (a drone with propellers on six arms) that weighed about 2kg to study changes in Llaca Glacier in the central Cordillera Blanca of the Peruvian Andes.

Llaca, one of more than 700 glaciers in the Cordillera Blanca, was chosen for both logistical and scientific reasons. It covers an area of about 4.68 square kilometers in Huascaran National Park and spans an altitudinal range of about 6000 to 4500 meters above sea level. Like other glaciers within the Cordillera Blanca, it has been retreating rapidly because of anthropogenic climate change.

The researchers hiked to the glacier to conduct surveys (Source: Oliver Wigmore).

To obtain footage, the researchers had to drive three hours on a winding, bumpy road from the nearest town, located about 10km away from the valley. “This was followed by a halfhour hike to the glacier,” Wigmore stated.

To overcome some of the challenges of working in a remote, high-altitude region, the drone was custom-built using parts bought directly from manufacturers. In this case, a base was bought from a manufacturer. “I modified it by making the arms longer, lightening it with carbon fiber parts, and adding features like a GPS, sensor systems, infrared and thermal cameras, and other parts required for mapping,” Wigmore shared.

Building their own drone allowed the researchers to repair it or replace parts when necessary, as sending it off to be repaired while in the field was not possible. It also allowed them to customize the drone to their needs.

A drone selfie taken by Wigmore, with the shadow of the drone in the bottom right corner (Source: Oliver Wigmore).

“No commercial manufacturers could promise that our equipment would work above an altitude of about 3000m, which is well below the glacier,” Wigmore said.

Using drones to study glaciers has advantages over conventional methods in terms of access to glaciers and spatial and temporal resolutions of data. These advantages have been further enhanced by hardware and software developments, which have made drones a relatively cheap, safe and accurate remote sensing method for studying glaciers at a finer scale. For example, Wigmore can build a UAV for about $4000, compared to the high cost of airplanes and satellites also used in remote sensing.

Wigmore and his team carried out aerial surveys of the glacier tongue (a long, narrow sheet of ice extended out from the end of the glacier) and the proglacial lake system (immediately beyond the margin of the glacier) in July 2014 and 2015. The drone was flown about 100 meters above the ice while hundreds of overlapping pictures were taken to provide 3-D images and depth perception.

High resolution (<5cm) Digital Elevation Models (DEMs) and orthomosaics (mosaics photographs that have been geometrically corrected to obtain a uniform scale) were produced, revealing highly heterogeneous patterns of change across the glacier and the lake. The data also revealed that about 156,000 cubic meters of ice were lost within the study period.

High resolution images showed rapid ice loss around exposed cliffs and surface ponds (Source: Wigmore and Mark, 2017).

The images revealed, for example, that the location of exposed cliffs and surface melt water ponds serve as primary controls on melt rates in the glacier tongue. Exposed cliffs lack the insulation of thick debris that are common on the glacier tongue, while ponds are less reflective than ice and absorb more solar radiation, causing higher melt rates.

The thickness of debris layers on the glacier constitute a secondary control. Thicker layers (often over 1m deep) provide insulation from solar radiation, while thinner layers increase the absorptivity of the surface to solar radiation.

The study also found that the upper section of the proglacial lake contains sections of glacier ice which are still melting. This suggests that the extent and depth of the lower section of the lake will increase as the ice continues to melt. This could increase the risk of GLOF, as expansion of the lake will bring it closer to the steep headwalls of the valley, which are potential locations for avalanche and rockfall debris.

Wigmore’s research is part of a series of larger projects still under publication that involve using drones to study glaciers, wetlands and proglacial meadows in the region. The results contribute to our understanding of hydro-social changes in the Cordillera Blanca, and how they can be managed.

Find out more about drone research here, or learn about Wigmore’s other research here.

Please follow, share and like us:

Share your thoughts