Using observational data to constrain CMIP6 model projection

Abstract:

The Arctic is one of the regions where the strongest climate change signals are observed on our planet. Projections of future atmosphere, ocean, and ice conditions in Arctic are important not only for the climate change studies, but also for stake holders and decision makers. In this paper we investigate how the most up-to-date climate models (CMIP6) simulate the present-day Arctic climate and what are their future projections under different emission scenarios. A comparison of the future projections with and without using observational record as a constrain in the present day simulations will be presented.

Compared to the observational records, the CMIP6 multi-model mean (MMM) overestimated the March sea-ice extent, and under estimated the September sea-ice trend in the last 20 years. For Ocean salinity, the MMM reproduce a falling sea-surface salinity trend in the Amerasian Basin, and show fresher Siberian shelves and a saltier Beaufort Gyre. Late-summer (September) sea ice stabilizes at more than half its present value under SSP1-2.6, while the Arctic Ocean’s sea-ice extent falls below the “ice-free” threshold of 1 million square kilometers by 2100 under the other scenarios. A dipole change pattern of the salinity (saltier in Eurasian Basin, and fresher in the Amerasian Basin is projected by the MMM. The warming in the Arctic will continue to be amplified under all scenarios, with the largest warming projected to happen in winter.