Challenges in Constraining Modeled Cloud Susceptibility to Aerosols Using Satellite Observations

In spite of decades of research, aerosol effects on clouds are still a major source of uncertainty in climate projections. Satellite retrievals are an important source for observationally based estimates of aerosol effects, and discrepancies between model and satellite estimates have led the climate modeling community to be concerned that clouds simulated from global climate models are overly sensitive to aerosol changes. This disagreement has driven modelers to change model formulations to increase agreement.

To reconcile modeled and observed aerosols, we have recently developed an aerosol lidar simulator that uses consistent definitions of aerosols and considers only the portion of the atmosphere sounded by the CALIPSO satellite. For the first time to our knowledge, a common algorithm is used in a global climate model and satellite observations to facilitate a consistent comparison of aerosol effects on clouds. We show that when the model and Earth are viewed through the same lens, discrepancies are greatly reduced because retrieved cloud reflectivity and lifetime are significantly less sensitive to aerosols than the real sensitivity. This study indicates that modeled aerosol-cloud interactions may be much closer to satellite-based estimates than commonly perceived, but the estimate may be significantly biased compared to the true situation, and suggests that traditional model evaluation without simulators could drive model development in the wrong direction. We highlight the importance of a common framework between climate models and real world when assessing aerosol indirect effects and calibrating aerosol and cloud processes in climate models.