Predicting tidewater glacier retreat

The Greenland Ice Sheet (GrIS) will be one of the largest contributors to 21st Century sea level rise, but quantifying the expected mass loss has proven difficult due to the challenge of predicting the retreat of tidewater outlet glaciers (i.e. those glaciers that drain directly into the sea). The rate of retreat of these glaciers can vary dramatically on short timescales (e.g. Howat et al, 2008), but it remains unclear whether these changes are driven by warming of the atmosphere, ocean or both (Carr et al, 2013). Improving understanding of the controls on tidewater glacier retreat is, therefore, vital if its contribution to the mass budget of the GrIS is to be predicted by models (e.g. Nick et al., 2013). However, the remote and dynamic nature of these environments makes the collection of field data extremely challenging, limiting progress in this area.

Recent research by Cowton et al. (2018) has used remote sensing as a method for studying a sample of tidewater glaciers in east Greenland. In an attempt to simplify predictions of tidewater glacier retreat, this research investigated the extent to which retreat can be explained by four basic variables: meltwater runoff, ocean temperature, and two simple parameterisations of 'ocean/atmosphere' forcing based on the combined influence of these two variables. The results demonstrate that the retreat of east Greenland's tidewater glaciers is best explained as a combined function of both oceanic and atmospheric warming. Furthermore, this research found that despite the complexity of tidewater glacier behaviour, over multi-year time scales a significant proportion of terminus position change can be explained as a simple function of this forcing (Cowton et al., 2018).

While this research provides an excellent starting point for the parameterisation of complex ice-ocean-atmosphere interactions, this research is limited in its scope, due to its small sample size (ten tidewater glaciers), located in a geographically constrained region of east Greenland. Additionally, the research used datasets which were temporally limited to a period of 20 years. Therefore, the research proposed here aims to take the principles established by Cowton et al. (2018) and apply them more broadly, both in geographical and temporal extent, in order to investigate the wider applicability of the findings. For an expanded sample of glaciers (covering all regions of Greenland), time series of terminus retreat will be created by digitising glacier terminus position in a range of freely-available satellite imagery (e.g. Landsat, Sentinel-1 and 2). These time series will be analysed in conjunction with temporally longer datasets of atmospheric and oceanic conditions (e.g. Noel et al, 2016). If simple relationships between ocean-atmosphere forcing and the retreat of many different glaciers can be established, this research would contribute significantly to the field of study, allowing simple parameterisations for tidewater glacier retreat that could play an important role in predicting the response of the GrIS to future climate warming.