Master Thesis Defense by André Emil Toft Jensen

Title: Remote sensing of Arctic surface albedo feedback

Arctic surface albedo feedback (SAF) amplifies global warming and drives Arctic amplification. While sea ice and snow cover retreat are known to alter surface albedo, the resulting forcing remains debated. SAF is calculated by mul- tiplying surface albedo sensitivity to global mean surface temperature (GMST) with the atmosphere’s radiative sensitivity to albedo change. Traditionally, this radiative sensitivity is estimated using climate model simulations of so-called radiative kernels. Here, a fully observational approach is taken with CLARA-A3 blue sky surface albedo data and remote sensing estimates of radiative sensitiv- ity. To this end, a simple one-layer atmosphere model was applied, improved, and compared to other radiative kernels. The radiative kernels are evaluated by comparing the surface albedo radiative forcing to observed energy fluxes and cloud properties. The basic one-layer atmosphere model and two radiative kernels derived from GCM simulations showed high sensitivities that couldn’t be explained. Then, an empirical kernel model is regressed onto 1979-2023 surface albedo data to address the error of using static radiative kernels. The relative error of using kernels based on 2001-2023 to evaluate the 1979-2023 sea ice forcing is estimated to be -6% and about one-fourth for land snow. The final remote sensing estimates of the 1979-2023 SAF were approximately 0.19W m−2 K−1, 0.08W m−2 K−1, and 0.27W m−2 K−1 for sea ice, land snow, and the cryosphere, respectively. Exact uncertainty estimates were not made. Analyzing 1979-2023, we find that sea ice SAF accelerated in the 2000s, while land snow SAF in the traditional sense has only been significant since the early 2010s. These results align with previous studies applying the radiative kernel technique to observations of sea ice but are significantly lower than observational land snow SAF assessments. I also raise questions about the validity of another assessment method, which assumes that the clear-sky and all-sky planetary albedo have a constant relation in time.

Supervisors: Aslak Grinsted, Gorm Dybkjær (DMI),  Jacob Høyer (DMI) and Jason Box (GEUS)
Censor: Sebastian Bjerregaard Simonsen (DTU)