Environment

The volume of the world’s glaciers, as estimated by a new AI model

Developed by Ca’ Foscari and the CNR, the model combines millions of data points and physical variables to reconstruct glacier thickness, thereby improving climate and water forecasts

 Adobe Stock

2' min read

Translated by AI
Versione italiana

2' min read

Translated by AI
Versione italiana

A new artificial intelligence model for studying glaciers. And to answer the question of how much ice they might contain. Around 150,000 cubic kilometres, equivalent to a 32.3-centimetre rise in mean sea level, should all the glaciers melt completely (excluding Antarctica and Greenland).

The study

This was established by a study coordinated by Ca’ Foscari University of Venice in collaboration with the CNR-ISP and published in *Scientific Data* (Nature). The model in question is IceBoost v2.0, a machine learning model developed by Niccolò Maffezzoli, a physicist and researcher at Ca’ Foscari University of Venice, affiliated with CNR-ISP.

Loading...

7 million measurements

The model, trained using over 7 million measurements, reconstructs the volume and distribution of the ice. A web app also allows users to visually explore data on all the Earth’s glaciers. The CNR emphasises that the model ‘combines this data with 26 physical and geometric variables (including terrain slope and curvature, ice velocity and temperature)’. ‘In this way,’ they explain, ‘it reconstructs, point by point, the thickness of the ice and therefore the volume of every glacier in the Randolph Glacier Inventory, the most up-to-date map of existing glaciers, excluding the polar ice caps’.

150,000 cubic kilometres

According to the new estimate, the planet’s glaciers contain around 150,000 cubic kilometres of ice, equivalent to a 32.3-centimetre rise in mean sea level, should all the glaciers (excluding Antarctica and Greenland) melt completely. ‘The overall figure is in line with the results of previous studies,’ the CNR emphasises in a statement. ‘However, IceBoost 2.0 is able to estimate the distribution of each glacier more realistically, resulting in estimates that are up to 40 per cent more accurate than field measurements.’ That’s not all:

“One of the most significant examples concerns the Geikie Plateau in eastern Greenland, where the glacier is up to 2 kilometres thick,” they argue. “In this area, IceBoost v2.0 estimates that there is almost twice as much ice as previously reported.”

The scientist

According to Niccolò Maffezzoli, ‘the distribution of ice thickness is a key variable for glaciological and climate models’. “To predict how glaciers will evolve by 2100 and their contribution to sea-level rise, we need to understand their current state in as much detail as possible,” the researcher emphasises. Researchers from the Glacier Model Intercomparison Project (GlacierMIP4), who are working on the forthcoming simulations to inform the IPCC about glacier evolution by 2100, will use IceBoost v2.0 as the sole representation of the current situation.”

Future campaigns

The new dataset also serves as a useful basis for guiding future measurement campaigns and planning scientific investments. Glaciers feed rivers, ecosystems, agriculture and local communities, supporting a total of 1.9 billion people across the globe. ‘A better understanding of the thickness and volume of the ice,’ the study emphasises, ‘helps to develop more reliable scenarios regarding the future availability of fresh water.’

Copyright reserved ©
Loading...

Brand connect

Loading...

Newsletter

Notizie e approfondimenti sugli avvenimenti politici, economici e finanziari.

Iscriviti