Past and Future High-resolution Global Glacier Mass Changes

Introduction

World-wide glaciers are losing mass which affects global sea-level, river runoff, freshwater influx to the oceans, glacier-related hazards, and landscape changes, with implications for human livelihoods and ecosystems. Hence, accurate estimates of past, current, and future glacier mass variations at a high temporal and spatial resolution are key to effective adaptation strategies.

Challenges in Current Models

However, previous mass-balance reconstructions and projections have relied on scarce observations with limited spatial and/or temporal resolution. Additionally, they have used overparameterized, insufficiently constrained, and highly simplified models, which were necessitated by the high computational costs incurred by the global scale.

GLACMASS Overview

GLACMASS will propel the current state-of-the-art of global-scale glacier reconstruction and projection forward in unprecedented ways. It will deliver a fundamentally novel and internally consistent physically-based modelling framework that draws, for the first time on a global scale, on both data assimilation and modern machine learning techniques facilitated by emerging global-scale glacier-related satellite-derived data.

Objectives of GLACMASS

The framework will be used to:

1. Reconstruct multi-decadal past glacier changes.
2. Make policy-relevant multi-century projections of mass and area changes of all >200,000 glaciers outside the ice sheets with unprecedented accuracy, spatiotemporal detail, and computational efficiency.
3. Nowcast present mass changes in a near-real-time fashion for selected regions.

Methodology

The model framework will fuse output from a novel physically-based glacier evolution model with all relevant observations available for each glacier. This includes:

• In-situ measurements
• Geodetic data
• Gravimetry-derived mass balances
• Snowlines and other observations

Thus, it will simultaneously exploit the untapped strengths of different types of observational data sets in an optimal manner.