An observational estimate of Arctic UV-absorbing aerosol direct radiative forcing on instantaneous and climatic scales

Abstract. Using co-located satellite observations from the Aqua Moderate Resolution Imaging Spectroradiometer, the Aqua Cloud and the Earth Radiant Energy System, the Special Sensor Microwave Imager/Sounder, and the Ozone Monitoring Instrument, we investigated changes in absorbing aerosol direct radiative forcing (ADRF) in the spring through fall Arctic from 2005–2020 through an observation-based method, assisted by a neural network for estimating aerosol-free sky top-of-atmosphere (TOA) radiative fluxes, and an innovative, Monte Carlo-based method for estimating uncertainties in derived ADRF values. This study suggests that Arctic ADRF is a strong function of observing conditions, and changes in Arctic sea ice concentration (SIC) and cloud properties introduce a complex scenario for estimating ADRF. For example, the TOA ADRF reverses sign from negative (cooling) to positive (warming) for SIC above 60 % for a region with a relatively cloud-free scene. ADRF trends over Arctic land surfaces are primarily negative. Strong negative ADRF trends of up to −4 W m−2 were found over northern Russia and northern Canada in the summer months. Both positive and negative ADRF trends were found over the Arctic Ocean in the boreal summer, though these trends are much weaker than the over-land trends. Positive ADRF trends in the Arctic Ocean north of northeastern Russia and northern Canada are as high as +1.0 W m−2 per study period. The trend results suggest that increasing amounts of absorbing aerosols in the Arctic have a cooling effect from TOA that could act to counter Arctic warming.