How do clouds and turbulence shape the northeasterly low-level jet of a summertime Arctic cyclone during its cold core transition?

Abstract. Synoptic-scale Arctic cyclones play a major role in the polar climate variability by bringing heat and humidity from lower latitudes. However, they have not been as intensively studied as mid-latitudes cyclones, and some questions about their development processes remain unresolved. The international THINICE campaign which occurred in August 2022 near Svalbard archipelago aimed at studying Arctic cyclones and in particular low-level jets (LLJs) along fronts that can favour sea ice breakup and contribute the heat and moisture meridional atmospheric transport. By comparing ERA5 reanalysis with radar observations from the campaign, we show that ERA5 reasonably captures the LLJs correctly, in particular several cold-conveyor-belt (CCB) jets, but tends to slightly underestimate their intensity and to position them not close enough to the surface. It is therefore essential to understand the processes that drive the structure and intensity of these LLJs among which diabaticand frictional processes associated with clouds and turbulence that is, processes that are fully parameterised in numerical weather prediction and climate models. The objective of the present work is to understand how clouds and turbulence shape an intense LLJ during the typical warm-core to cold-core transition of a well-observed Arctic cyclone during THINICE. The analysis is conducted using the ERA5 analyses and short-term forecasts. It is based on the computation of backward Lagrangian trajectories initialised in a narrow band of high potential vorticity (PV) associated with the northeasterly LLJ. A PV budgetis performed in which all the diabatic – mostly associated with cloud condensation and radiative effects – and frictional PV tendency terms are computed and investigated. Results show that both diabatic and frictional terms reinforce the potential vorticity gradient across the jet and therefore its intensity when the system is baroclinic. Under such a baroclinic growth phase, the northeasterly LLJ is well identified as a CCB jet along the bent-back warm front. However, when the system is not baroclinic anymore and reaches a cold-core structure, friction generates much less PV and has almost no effect. Radiative cooling above boundary layer clouds no longer generates PV either, unlike during the baroclinic phase, and even slightly destroys it. Heating from condensation inside thick frontal clouds generate the same amount of PV in both phases (baroclinic and cold-core phases). Overall there is less PV generation in the second phase than in the first phase, which overall leads to a less intense LLJ.