Standardising pack ice measurements to better predict climate



At UCL’s Georges Lemaître Centre for Earth and Climate Research (TECLIM), two researchers are passionate about polar region climate variations: François Massonnet and Hugues Goosse. ‘These regions fascinate me’, Dr Massonnet says. ‘They’re the sites of the last remaining almost virgin human activities on our planet. Through our human footprint, however, we have managed to upset these ecosystems, which are a priori fully protected.’

For many years, climate studies have sought to understand these upheavals over very long periods because doing so could help predict climate change more accurately. Some of these studies cover several millions of years in the past and into the future. ‘Hugues Goosse and I are more specifically interested in the climatic variations in these polar regions over shorter periods, of a few dozen years. From the end of the 1970s to the end of the present century, we analysed the evolution of the pack ice and all the models that have been created for this period. One observation is that no two models are alike: all agree that the pack ice melts but not on how fast.’

Too many differences among forecasts

Pack ice, the thin layer of ice that floats and naturally forms in the ocean, is to be distinguished from the thick ice cap. For many years, many studies have been in agreement that this area of the globe is connected to climate change. ‘In satellite and field studies, it’s clear that there’s a significant withdrawal of pack ice’, Dr Massonnet confirms. ‘All studies attest to this. But none of them agree on melting speed.’ The great interest of the study conducted by Dr Massonnet and Prof. Goosse is to have identified the origin of the disparities in pack ice melting forecasts: they’re a consequence of how the modellers estimate its thickness to begin with. At present, pack ice can be measured only by satellite. From space, satellites send signals to the heart of the pack ice. Once impacted, the surface returns the signals to the satellite, which calculates ice thickness via the distance travelled. But this transmission is disturbed by various elements, ‘noise’ consisting of, for example, snow on the pack ice and atmospheric turbulence. Measuring pack ice thickness is therefore a challenge. Scientists talk about uncertainty ranging from 50 to 100%. So for every meter calculated, the error can be 50 centimetres to an entire meter – the entire measure. So how to choose the ‘right’ models, with the right ice thickness? ‘We can’t be sure because the time scale is too short: we’ve only been collecting data on the evolution of pack ice for 15 years. But in climate terms it takes at least 30 years to ensure certain results.’ Conclusion: such a measure is impossible today.

Call for space observation programmes

‘So we’re launching a genuine call for space observation programmes to deploy large-scale observation systems and constrain model prediction formats to better predict pack ice melting’, Dr Massonnet says. ‘We can’t get 30 years of experience all at once, but we can at least hope for more accurate observations.’ The research community might realise that hope. ‘Five years ago, the scientific communities of model development were separate and didn’t speak to each other very much’, Dr Massonnet explains. ‘Since then, more and more conferences have brought together modellers and climate observers, like Hugues and me, we both wear multiple hats, and real joint work is taking place. We realised this is the key to future research.’

Marie Dumas


François Massonnet is an F.R.S.-FNRS postdoctoral researcher and a member of the Georges Lemaître Centre for Earth and Climate Research (TECLIM, Earth and Life Institute). With Hugues Goosse, he published the results of a study on pack ice evolution and the different ways of calculating its progressive decrease.

Why focus on the polar regions?

In addition to passion, there are three main reasons that justify our interest in these areas of the globe. The first is that they’re on the front lines of climate change. The Arctic has warmed much more quickly than the rest of the world and so have several parts of Antarctica. To use a fairly common expression in our environment, these regions are a bit like the ’canary in a coal mine’, that is, they’re early indicators of climate change. Another reason is that we have more and more evidence that what’s happening in the polar regions isn’t destined to be confined to them. These are ‘teleconnected’ regions: they’re linked to other, physically distant regions which, a priori, should not be in direct interaction but in any case interact in one way or another. This is the case, for example, in the Pacific Basin with the El Niño phenomenon and its repercussions in the Atlantic. We observed an identical effect between the Arctic and Europe. Several studies have also shown that these strong warmings in the Arctic could induce variations of cold or heat all over the globe. These regions can therefore be indicators of our future climate.

The evolution of this area of the world could lead to many changes in the future.

This is the third reason that guides this study. It’s also a more pragmatic reason. These areas have become more inhabitable. Although they remain cold in winter and ice-cold most of the time, we have observed a significant withdrawal of pack ice during summer, allowing boats to pass through for the first time. This opens up opportunities for many economic actors and redrawing maps at the geopolitical level. It’s a region rich in possibilities and yet to be discovered.





A glance at François Massonnet et Hugues Goosse’s bio

François Massonnet is an F.R.S.-FNRS postdoctoral researcher and a member of the Georges Lemaître Centre for Earth and Climate Research (TECLIM, Earth and Life Institute). After graduating as a civil engineer in applied mathematics at UCL in 2009, he proposed in his PhD thesis innovative methods to reconstruct and predict the evolution of the Arctic and Antarctic pack ice through the joint use of climate models and satellite and field observations. From 2014 to 2016, he was a postdoctoral fellow at the Barcelona Super Computing Center in Spain as part of a world-renowned team in climate forecasting for periods ranging from a few months to a few years. In particular, he has gained recognised expertise in the areas of assessment and verification of climate models and forecasts. Through this experience, he contributed in 2013 to the writing of the fifth report of the Intergovernmental Panel on Climate Change (IPCC) and is currently involved in two Horizon 2020 European projects. More information: et @FMassonnet (Twitter).

Hugues Goosse is a professor at UCL where he teaches climatology and related disciplines, and a research director at the Belgian Fund for Scientific Research (F.R.S - FNRS). His research focuses on the development of climate models, comparisons between model results and different types of observations, and the application of these models to study past and future climate change. He concentrates on both natural variations and climate change induced by human activities. More specifically, this recent work focuses on the interactions between sea ice and the ocean in the Southern Ocean, climate change during the last millennia and methods of data assimilation in palaeoclimatology.

Published on July 12, 2018