Constraining Model Representations of Shallow Convective Mixing and Shallow Cumulus Physics with Observations of Stable Water Isotopes

Shallow convective mixing between the boundary layer and the free troposphere is a key physical process determining boundary layer moisture and hence low-cloud formation and properties. As the physics of low-cloud feedbacks contributes extensively to uncertainty in estimated climate sensitivity, the representation of shallow convective mixing in climate models is critical to constrain. However, it is hard to compare directly to observations due to the difficulty in measuring convective mixing itself at the relevant spatial and temporal scales. As a novel approach, stable water vapor isotope profiles from satellite retrievals have been demonstrated to be effective tracers of convective mixing in shallow cumulus regions (i.e. southern Indian Ocean and Pacific gyres). With this established connection, we constrain the representation of these low clouds in an isotope-enabled global climate model (NCAR CESM). With a perturbed parameter experiment in iCAM5 (CESM1), we highlight the influence of the in-cloud autoconversion rate, re-evaporation of hydrometeors, and lateral entrainment of dry free-tropospheric air. We find, however, that the largest changes in the low cloud-shallow mixing relationship emerge when transitioning from one generation of the shallow convection/PBL scheme to the next. In demonstrating that the newer scheme implemented in iCAM6 fits better with available observations – including both NASA satellite retrievals of isotopic estimates of shallow convective mixing and shallow cumulus properties – compared to iCAM5, we identify the key components generating an improved fit. We show that using water isotopes as a physical constraint on the relationships between mixing and shallow cumulus allows us to develop a new benchmark for evaluating both the climatology and feedbacks associated with low clouds in a global model.