Research Interests

Glacier Dynamics

Glaciers are retreating worldwide as a result of global climate change and understanding how they will evolve is a major challenge. Central to this problem is a lack of understanding of the key processes that underpin these changes, such as glacier calving and its influencing factors. I am interested in how glaciers behave over a variety of spatial and temporal scales, ranging from localised calving events to regional scale changes in ice velocity and dynamics. The process of ice movement is complex and is influenced by a multitude of factors. Disentangling these factors is difficult and requires a mixture of observations and process-based modelling. My research focusses on generating accurate data sets at sufficient spatial and temporal resolution to understand these changes. I focus predominantly on the Arctic and have conducted research in Scotland, Svalbard, Greenland and the Canadian Arctic.

Close-Range Remote Sensing

I have a strong interest in developing Radar techniques for Glaciology. In my research, I am developing the method of ground-based millimetre wave radar (94 GHz) for glacier remote sensing. Millimetre waves are at the high frequency end of microwave techniques and utilise the benefits of high-resolution optical instruments and the all-weather capabilities of low frequency radar systems. Further, ground-based sensors can measure glacier processes at the scale of hours to days, elucidating changes on spatial and temporal scales unresolvable by satellite techniques. This exciting new research will be applied to a number of glaciological problems (see here) and will enhance the suite of Remote Sensing instruments able to monitor glacier processes. Central to this project is understand the radar backscatter characteristics of glacier ice at 94 GHz and preliminary work was present at the EGU General Assembly (see here).

The AVTIS2 millimetre-wave radar being developed for glacier remote sensing. Photo from W.D. Harcourt

I was fortunate enough to attend the 2019 Innsbruck Summer School of Alpine Research where I learnt the theory and application of a number of close range Remote Sensing techniques: UAV Photogrammetry, Terrestrial Photogrammetry, Lidar and Terrestrial Laser Scanning (TLS). TLS is an optical surveying instrument operating in the visible and infrared frequency range. It emits light and measures the total travel time to determine the range to a target. We used a Riegl TLS system to monitor landslide dynamics near Obergurgl, Austria. One of the key processing steps during TLS post-processing is to classify the point cloud into ground and non-ground points. We tested three algorithms to do this and the results present at the XXIV ISPRS congress (see here).

Terrestrial Laser Scanner at an active landslide, Austria.

Satellite Remote Sensing

The satellite era of Earth Observation (1960s) has exploded in recent years with the an exponential rise in the number of satellite observing the Earth’s surface. Because of this, we are living through an age of ‘Big Data’ which demands us to explore more complex ways of analysis this detailed history of the modern Earth. I have used an array of satellite data (optical, radar, SAR, altimetry etc.) to understand how the planet is evolving under the constant strain of climate change. For instance, I have derived a time series of glacier velocity for the fastest-flowing glacier in the Canadian Arctic, developed techniques to map supraglacial drainage pathways across the Greenland Ice Sheet and quantified elevation changes of outlet glaciers from satellite altimeters.

Sentinel-2 time series of Helheim Glacier, SE Greenland (created using the Sentinel Hub EO Browser).

I have also applied satellite data analysis to understand tropical ecosystems in inland coastal waters. Deriving spatial coverage estimates of seagrass is difficult due to their growth in water and their seasonal growth and decay. In order to aid spatial coverage estimates in East Africa, where studies on tropical coastal ecosystems are severely lacking, we analysed Landsat and Sentinel-2 imagery to understand 30-year changes in seagrass coverage. We found a decline in seagrass coverage across the coast of Kenya that accelerated beyond the 1980’s. Future studies should look to integrate these data sets with other spatial data on coastal population, land use changes and ocean temperatures in order to isolate the drivers of these changes. We also used underwater GoPro images (see below) to act as validation for our seagrass maps.