Diagnostic Relationships in Observed and Modeled DYNAMO convection

The Madden-Julian Oscillation (MJO) is poorly represented in climate models. The poor representation of the MJO is related to inadequate convective parameterizations, due to lack of understanding of how local large scale processes influence local convection. To better understand how large scale processes influence local convection, we compare diagnostic relationships between observations and simulations of DYNAMO convection. We compare diagnostic relationships between rain rate, saturation fraction (column relative humidity), instability index (a measure of column air stability), and normalized gross moist stability (NGMS, ratio of export of a quantity conserved in moist adiabatic processes, and moisture import).

Observed diagnostics were calculated from the Colorado State DYNAMO gridded data sets. Modeled diagnostics are computed by a cloud resolving model utilizing the weak temperature gradient (WTG) approximation. The WTG approximation is based on observations that gravity waves quickly redistribute heating anomalies leaving weak temperature gradients in the tropics. In the model, the WTG approximation is implemented by adiabatically displacing air mass; vertical WTG mass transport is proportional to the temperature anomaly caused by latent heat release due to precipitation and radiative cooling. In WTG simulations, the model is forced by observed potential temperature anomalies, mixing ratio anomalies, sea surface temperature variations, and wind speed variations.

The model successfully reproduces the observed relationships between rain rate, saturation fraction, and instability index. Rain rate is a monotonically increasing function of saturation fraction and atmospheric stability. Diagnostic relationships between NGMS, rain rate, saturation fraction, and instability index, peak around a critical value of NGMS. The critical value of NGMS differs slightly for the observations and for the model, possibly due to the fact that the model is run on a two dimensional domain. The physical origin of the critical NGMS value is still an open question, and it might be related to mechanisms governing MJO formation.