Digital Twinning

GreenlandTwin Project

The GreenlandTwin project will develop a first-of-its-kind, digital twin of the Greenland Ice Sheet (GrIS) by assimilating data from spaceborne, airborne and ground-based sensors using machine learning techniques to develop a virtual representation of 21st Century ice sheet change. It will represent time-varying changes in the configuration of the ice sheet from a purely data perspective, offering a unique lens through which to explore past, present, and future ice sheet dynamics. This project is ambitious in scope and will deliver a step-change in our understanding of ice sheet change through observations by making full use of the large volume of data now acquired from spaceborne, airborne and ground-based sensors. This digital twin will combine expertise from numerous disciplines (e.g. glaciology, ecology, climatology, oceanography, sociology) and with varied interests (e.g. model application, model development). This project started in 2023 and further updates will be posted on this page.

Data Fusion

Current data products that have been developed over the Greenland Ice sheet range is coverage and resolution, as well type (e.g. point clouds, 2D images, coverage maps). Therefore, to develop a digital twin product an effect methodology to combine these data sets is required. We are now developing multi-model, multi-resolution techniques to fuse these data products using machine learning methodologies.

Ice Sheet Forecasting

The primary aim of the digital twin is to develop the capability to forecast ice sheet changes into the future and back in time. This data-driven perspective on future changes to the Greenland Ice Sheet will provide a unique perspective on ice sheet change and complement theoretical models that have been developed based mostly on ice sheet physics. The forecasting capabilities will also expose the impact of different processes on ice sheet changes and also highlight gaps in our observational capabilities, hence opening up new opportunities to develop new data products that can be used for forecasting purposes.