Assessing Deep Convection's Moisture Sensitivity Function Using DYNAMO Observations

The dependence of convective heating on moisture is fundamental for understanding its larger-scale impacts and coupling, and knowing it may also offer clues for a deeper physical understanding of the convection itself. This dependence, within a convecting basic state, may be cast as convection’s response to small perturbations in the large-scale moisture field. For small enough perturbations, the dependence are formulated as a matrix, we may call the convection response functions (CRF). First-guess CRFs have been derived by Kuang (2010,2012) for a few different geometries and base-state convection strengths and profiles, through the use of carefully controlled simulations from a Cloud system resolving model (CSRM). This study scrutinizes the skill of these CSRM-constructed CRFs in two ways: (1) by using CRF information diagnostically to interpret observations from the DYNAMO field campaign, and (2) by using it as an inline representation of convection in a simplified global model. Part 1, the diagnostic work, includes converting the CRF into predictions in the space of field observables. Specifically, divergence profiles are retrieved from Doppler radar data on an hourly basis, which is a measure of convection’s heating strength, while humidity profiles are obtained from microwave radiometry as well as balloon soundings, and CRFs can be cast as (partial) predictions about some associations we might see in these data. The second effort involves a ‘dry’ (without convection scheme) primitive equation solver with calibrated, time-independent forcings to keep it in a realistic Northern autumn season basic flow regime, with CRFs slotted in as a linear anomaly version of convection. The matrix can be edited for experiments, and the skill of hindcasts for a few DYNAMO cases to define better versus worse matrices, to give an indirect but ultimately relevant view of the dependence of convective heating on moisture.