ESA CCI Research Fellowship: Stephen Chuter

Project: Fingerprinting Approach to close Regional Sea Level Budgets using ESA-CCI

Fingerprinting Approach to close Regional Sea Level Budgets using ESA-CCI (FACTORS)

Executive Summary

This fellowship has presented a significant step forward in quantifying and modelling the “Sea Level Fingerprints” phenomena. It has developed an agnostic framework for integrating EO products from a diverse array of observational platforms into a state-of-the-art sea level equation solver, focusing on appropriate error characterisation. This project therefore lays the groundwork for regional sea level budget closure assessment, the next goal after the successful closure of the sea level budget in the ESA CCI Sea Level Budget Closure project. It also demonstrates the necessity of large scale EO based assessments of key planetary variables enabled under the European Space Agency's Climate Change Initiative (CCI).

Sea level rise (SLR) represents one of the most significant global socio-economic issues of the 21st Century. Whilst global mean sea level is an Essential Climate Variable (ECV) for understanding how the Earth system responds to climate change, assessing socio-economic impacts and the need for adaptation strategies are driven by local sea level change rates, which can deviate significantly from the global mean.

A prominent driver of these local variations are the effects of ice sheet, glacier, and terrestrial water storage mass changes on the Earth’s gravitational field, rotation, and solid-Earth response. For many years, we have had the theoretical framework to calculate this component of regional sea level variability, referred to as “Sea Level Fingerprints” (SLFs). However, until recently, with the advent of satellite Earth Observation (EO) and its ability to pinpoint the spatial and temporal locations of ice mass loss, it was difficult accurately quantify this phenomenon. With land ice projected to be an increasingly large contributor to SLR over the next century, addressing this research gap is of imperative importance.

We utilised state-of-the-art EO and model derived land ice mass change ECVs from across the European Space Agency's Climate Change Initiative (CCI), to produce a monthly record of SLF fingerprints from 1992-2017. The diverse spatial and temporal variability in these EO products (e.g., high-resolution altimetry and coarse-resolution gravimetry) presented challenges in the appropriate integration and robust propagation of their uncertainties when calculating the SLFs using the sea level equation. To address this, we developed a novel statistical (Gaussian Process) sampling scheme, which produced hundreds of realisations of land ice mass change for each EO dataset that fully represented their original uncertainties. This framework provides a basis for future studies which require the integration of non-uniform EO products with differing error characteristics. 

This creation of hundreds of EO mass load realisations required a new way of solving the sea level equation, which is typically a computationally expensive exercise. We utilised and developed the Ice Sheet and Sea Level Model (ISSM), with its unique finite element mesh (FEM) parameterisation, to overcome this barrier. The FEM allowed for tailored high-resolution modelling in the regions of the largest mass loading change, such as Pine Island Glacier and Jakobshaven Isbrae, with coarser resolution over non-glaciated terrestrial landmasses. This setup, coupled with substantial development work in parallelisation and deployment to HPC architecture, enabled thousands of monthly SLF model realisations, which, when combined, has produced the most comprehensive understanding of this phenomenon to date. 

Our SLF fingerprints have highlighted key trends in this relative sea level (RSL) component, including substantial negative trends in the proximity of ice sheets (termed the ‘near-field’ effect), with the largest positive RSL trends in the equatorial regions. The novel monthly temporal resolution of our product enables new insights into SLF patterns, such as the visibility of seasonal variability driven by Greenland Ice Sheet, including step changes caused by the 2013 summer melt event. Additionally, negative RSL trends around the coast of East Antarctica are evident in the latter decade of the 1992-2017 period, reflecting the recent increases in ice mass loss in this region. A unique feature of this work, enabled by the individual ECV components of the ESA CCI, is for the impact of each land ice component (Greenland, Antarctica, and Glaciers) to be assessed individually as well as looking at the combined signal. 

This fellowship has presented a significant step forward in quantifying and modelling SLF phenomena. It has developed an agnostic framework for integrating EO products from a diverse array of observational platforms into a state-of-the-art sea level equation solver, focusing on appropriate error characterisation. This project therefore lays the groundwork for regional sea level budget closure assessment, the next goal after the successful closure of the sea level budget in the ESA CCI Sea Level Budget Closure project. It also demonstrates the necessity of large scale EO based assessments of key planetary variables enabled under the ESA CCI.

Publications

Host Institution: University of Bristol

Research Fellow: Stephen Chuter