Abstract

A significant uncertainty in our forecasts of Arctic climate is the Arctic cloud response to sea-ice retreat. This uncertainty derives from multiple factors, most significantly, from our lack of understanding of the source of high-latitude aerosol (to which Arctic clouds are peculiarly sensitive). It has been suggested that reduction in sea-ice will increase local emissions of aerosol increasing Arctic cloud albedo through the 1st (>albedo) and 2nd (>lifetime) indirect effects. Potentially Arctic cloud brightening may mitigate the ice-albedo feedback to some (unknown) degree. However, accurately predicting the response of Arctic aerosol and cloud to sea-ice retreat is predicated on our understanding of present-day aerosol sources and processes. Observations suggest that Arctic aerosol trends are driven by a combination of primary and secondary precursor gas emissions, and indicate a complex coupled system involving ice, ocean, atmospheric and biological processes. It is therefore no surprise that models struggle to reproduce observed Arctic aerosol and clouds. Historically, modellers have ‘filled the gap’ in modelled aerosol with local emission fluxes or processes which are ‘tuned’ to observations (varied until the model looks ‘right’). This tuning process is key to producing skilled models (akin to calibrating an instrument) however in a system of so many unknowns there is a danger that tuned emissions may be a self-fulfilling prophecy leading to a flawed understanding of Arctic aerosol and cloud.  Here, I will discuss the urgent problem with Arctic clouds in global climate model, introduce novel modelling methods which could inform observations and call for a combined use of modelling and observational techniques across diverse fields to (among other issues) understand the likely significant (but unconstrained) role of biological processes (from primary marine organics to ammonia emissions from bird colonies).