The onset of melt in the Arctic happens every year but the date can be very different from year to year. So, what really happens when Arctic sea ice starts melting in spring? Does it happen fast, are there any slower underlying or preconditioning processes and how does the ocean/sea-ice/atmosphere system interact through the time before and during this? The ARTofMELT was a unique expedition on icebreaker Oden in May and June 2023, in the sea ice of the northern Fram Strait, to explore the impacts of atmospheric warm-air intrusions, so-called atmospheric rivers, on the start of the melt season; hence the name, Atmospheric Rivers and The onset of Arctic MELT - ARTofMELT.
The early deployment of ARTofMELT was motivated by the need to better understand what atmosphere, cryosphere and ocean processes that govern the timing and character of melt onset. While the start of the autumn freeze is relatively well documented from scientific observations, the spring onset of the melt is not, due to very few science expeditions during winter and spring. And the reason is obvious; the difficult sea ice to brake and hence to navigate. ARTofMELT daringly targeted this period, taking interdisciplinary observations covering a vertical column from the upper ocean, through the sea ice and to the top of the atmosphere.
This was the second science workshop, roughly two years after the expedition took place, to review the progress in the different research groups with publishing datasets, scientific analysis, paper writing and collaboration. Like the first workshop, held a year earlier, it was supported by the the Atmospheric Working Group of the International Arctic Science Committee (IASC/AWG), Stockholm University’s International Meteorological Institute (IMI) and the Swedish Polar Research Secretariat (SPRS). Specifically, five PhD-students benefitted from Early Career Researchers travel grants from IASC/AWG. While the first workshop was held the week after EGU Annual Assembly, and a successful ARTofMELT-session, this workshop was held the week before the EGU; the back-to-back organization helped minimize overseas travel. Still the second workshop had a fairly large on-line participation; 18 of the 50:ish registered participants followed the workshop on Zoom.
A general idea in the program design was to merge modeling and observations, hence all sessions had a mix of modeling and observations discussed; each session also ended with O(30 min) discussion. One major purpose was to link the observations of meteorology, aerosols and aerosol precursors, atmospheric chemistry with ice characteristics and upper ocean structure across different scales from the different groups. These include observations directly from the icebreaker Oden, from airborne platforms launched from Oden such as the tethered Helikite and a helicopter-borne Helipod. Also, first results from detailed and improved atmospheric model developments were presented. The first day started after lunch and focused on meteorology, while the second day started with sea ice and ocean and ended with aerosols and atmospheric chemistry. At the end of the second day we watched the Stockholm University documentary movie and also welcomed ARTofMELT’s expedition artist, Ida Rödén that showcased her book on the expedition; this was followed by a hosted dinner. On the third day the workshop ended after lunch; before this time devoted to the future: the building of the ARTofMELT data collection at the Bolin Centre Database and future publications; especially a synopsis for an expedition overview paper was discussed. Tasks were assigned and the paper has now started to come together. Below are some science highlights:
The most intriguing results that were discussed was the progression of the system up to the melt. While the melt actually happened rather fast on 10 June 2023, closer inspection of different observations, indicate thresholds being passed in different ways on different dates preceding the melt onset. Generally speaking, more protracted but slower changes appeared under the ice already weeks prior to the melt onset, but after two significant atmospheric events; one brief melt period of less than one day followed by a storm. Notable among the changes in the upper ocean is the onset of a algal bloom first detected about two weeks prior to the surface melt onset. Changes in the isotopic mix of the water vapor observations indicate a shift after the first temporary melt but before the storm, while the characters of both aerosol size distributions and the surface energy budget diurnal cycle changed also about two weeks before the atmospheric river that triggered the melt onset arrived in the morning of 10 June.
Analysis of the energy budget also illustrates the role of low-level clouds. While cloud free conditions have the highest peaks of energy flux toward the surface, dominated by the solar radiation, it also comes with the largest surface-energy loss in the longwave portion of the electromagnetic spectra. It turns out that integrated over time, cloudy conditions provide a larger net influx to the surface. While reducing solar radiation, low clouds also cancel the longwave energy loss for a slightly positive net flux. Therefore. adding an influx of warm air can tip the system into melt by a downward turbulent sensible heat flux. Once the melt occurred in earnest, the characteristics of the surface changed fast, in less than one day and melt ponds started appearing on the second day. Unfortunately, ARTofMELT had to end on schedule only a few days into the melt period. Subsequent analysis of satellite data indicates that the 2023 melt onset date was about two weeks later than expected from climatology.
Highlights
- While the surface melt onset was fast, components the Earth system, below, in and above the ice, underwent changes, some gradual and some abrupt, through about three weeks prior to the final onset date.
- Low-level clouds play a decisive role in precondition the surface for melt triggered by an atmospheric river. Although clouds blck some solar radiation, the trapping of longwave radiation below clouds more than outweighs this loss and the surface energy budget is more often positive in clouds conditions than in clear.
- Once the surface melt has started in earnest the surface characteristics changes fast and melt ponds on top of the ice were observed in two days after the first zero-crossing of the near-surface air temperature.
Additional Material:
- https://www.polar.se/en/expeditions/previous-expeditions/arctic/artofmelt-2023/
- https://www.su.se/english/research/research-projects/artofmelt-atmospheric-rivers-and-the-onset-of-sea-ice-melt-2023
- https://www.su.se/english/research/current-research/artofmelt2023-expedition-studies-the-arrival-of-summer-in-the-arctic-1.646675
- https://www.su.se/department-of-geological-sciences/research/artofmelt-wp-10-insides-by-julia-muchowski-1.657198
- https://www.su.se/meteorologiska-institutionen/nyheter/artofmelt-en-unik-expedition-i-arktis-1.700634.
Date and Location
23-25 April 2025 I Stockholm, Sweden
IASC Working Group funding the Project:
AWG
Project Leaders:
Michael Tjernström (Stockholm University, Sweden)
Paul Zieger (Stockholm University, Sweden)
Year funded by IASC:
2024