Modeling the convective response to changes in the thermodynamic environment using the weak temperature gradient approximation: on the treatment of moisture

The weak temperature gradient (WTG) approximation is gaining broader use as a tool for investigating the response of convection to different thermodynamic environments. Recent work has incorporated observations from field programs in WTG simulations to gain insight into mechanisms controlling the MJO, cyclogenesis, and even diurnal variability over open oceans. While many of these studies demonstrate considerable skill in simulating observed precipitation rates, differences in implementation of the WTG approximation applied to complex time-dependent reference profiles may obscure the attribution of mechanisms, making it difficult to interpret the results. A systematic study with prescribed changes in reference environments which compares several choices in the implementation of WTG would quantify the effects of such choices and improve interpretation.

In the WTG approximation, reference profiles of potential temperature and moisture are specified to represent conditions outside the model domain. The domain mean potential temperature in the model is relaxed to the reference profile using a parameterized vertical velocity that counteracts diabatic heating. This vertical velocity vertically advects moisture which alters the vertical mass flux profile and influences the precipitation rate. Moisture external to the model domain enters the domain via 1) lateral entrainment due to an independent enforcement of mass continuity associated with the WTG velocity, or 2) by relaxation to the reference moisture profile. Lateral entrainment corresponds to divergent circulations associated with domain-wide ascent or descent. The moisture relaxation accounts for non-divergent advection of moisture associated with large scale rotation.

We perform a series of WTG experiments using either lateral entrainment or moisture relaxation in different thermodynamic environments. Different environments are represented by the reference profile, which is the mean radiative convective equilibrium state perturbed by 1) increasing or decreasing atmospheric stability, 2) moistening or drying the reference environment, or 3) a combination of the two. We then compare the modeled precipitation rate, the vertical mass flux profiles, and the gross moist stability--a diagnostic quantity related to precipitation efficiency--for the different thermodynamic environments using various moisture treatments. We find that the treatment of moisture can influence the properties of convection in different environments. Quantifying these differences will lead to improvements in interpreting simulated convection using the WTG approximation.