| Summary: | Mass losses from the Antarctic Ice Sheet are the largest uncertainty in predicting sea level rise. The role of surface meltwater in determining Antarctic sea level contribution is poorly understood, despite its potential to impact ice sheet mass balance via a number of mechanisms and feedbacks. This thesis uses novel methodologies, implemented within Google Earth Engine, to study the distribution and evolution of Antarctic surface meltwater at unprecedented spatial and temporal scales. An automated method for mapping surface meltwater is firstly developed and applied to the Amery Ice Shelf, and then at a continent-wide scale. The method overcomes previous challenges relating to data processing capability and variability in optical satellite image visibility, enabling the production of robust, continuous time series of surface meltwater area. The first continent-wide dataset of surface meltwater, spanning 2006 to 2021, reveals a significant increasing trend in surface meltwater area across East Antarctica since 2006. For the first time, clear links between surface meltwater area and large-scale modes of atmospheric variability are demonstrated, although these cannot explain the increasing trend observed in East Antarctica. Instead, this increase is likely a consequence of ice surfaces becoming more favourable to meltwater ponding. This thesis also explores novel techniques for improving quantification of surface meltwater volumes. A multi-satellite approach is applied to improve the temporal resolution of volume data, revealing a variety of mechanisms via which supraglacial lakes evolve at the Amery Ice Shelf. Combined automated and manual analysis additionally reveals the importance of melt-albedo feedbacks in controlling the spatial distribution of meltwater in the region. The complexity of meltwater processes motivates future studies to explore drivers and impacts of meltwater at a regional scale, including assessments of ice shelf vulnerability. Furthermore, the greater abundance of surface meltwater on ...
|
|---|