The Arctic is warming at two to three times as fast as the rest of the Earth, and it is therefore a crucial area of study for atmospheric scientists. However, the logistical difficulty of leading measure campaigns at high latitudes means that some key boundary-layer processes are still poorly understood. This thesis aimed to gain insight on two characteristics of the Arctic boundary-layer (clouds and surface based temperature inversions) and to determine their impact on the surface energy balance through a combination of novel measurements and modelling.

First, a novel statistic of cloud frequency and characteristics over the Arctic sea-ice was derived from a set of 1777 lidar profiles obtained during the 5-year Ice, Atmosphere, Ocean Observation Systems (IAOOS) campaign. Clouds were found to occur more than 85% of the time from May to October and single cloud layers were optically and geometrically thickest in October, possibly linked to moisture intrusions in autumn. Total cloud radiative forcing over a typical summer cycle was estimated to be negative for optically thin clouds, but positive for optically thick clouds.

Second, the impact of wind speeds on the development of surface based temperature inversions (SBI) in the continental Arctic was investigated. The analysis of measurements from the pre-ALPACA winter 2019 campaign that took place in Fairbanks, Alaska, showed that a local, likely topographically driven flow developed under anticyclonic conditions. This flow inhibited the development of strong SBIs by sustaining significant turbulence even under very strong radiative cooling. A transitional wind speed between weakly and strongly stable regimes was evidenced; this was coherent with the predictions of Minimum Wind speed for Sustainable Turbulence (MWST) theory. The modelling of clear-sky surface layer temperature inversions and their dependence on wind speed was then studied, with a focus on forest areas. A 2-layer analytical model of the vegetated surface layer was developed. This model exhibited a slower decrease of the SBI strength with wind speed compared to a 1-layer model, which was shown to be coherent with observations at an Ameriflux site close to Fairbanks. These models were then compared to two WRF (Weather Research and Forecasting) surface layer schemes, which were found to place excessive limits on the turbulence, preventing the development of large temperature gradients. The Arctic boundary-layer has become an active field of research in recent years. In this context, modelling advances and numerous planned campaigns open many perspectives for furthering the work presented in this thesis.

En présentiel : Sorbonne Université – 4 place Jussieu – Paris 5e – Salle de conférence de l’UFR 918, 46-56, 2ème étage.

En distanciel : https://us02web.zoom.us/j/89924696075?pwd=TjF0MnB6NllVY3Q4d1F1N1l5NnBrZz09

Meeting ID: 899 2469 6075 – Passcode: 573673

Solène Turquety – LATMOS (Présidente du Jury)
Julia Schmale – Ecole Polytechnique Fédérale de Lausanne (Rapportrice)
Timo Vihma – Finnish Meteorological Institute (Rapporteur)
Etienne Vignon – LMD (Examinateur)
Elsa Dieudonné – Université du Littoral Côte d’Opale (Examinatrice)
Javier Fochesatto – University of Alaska Fairbanks (Examinateur)
François Ravetta – LATMOS (Directeur de thèse)
Jean-Christophe Raut – LATMOS (Co-directeur de thèse)