The moist isentropic circulation from a dry isentropic perspective

In recent years, it has been shown that by averaging on isentropic surfaces it is possible to quantify the role of midlatitude eddies within the meridional overturning circulation. While this leads to realistic eddies mass fluxes, it does not generally account for poleward moisture fluxes. These moisture fluxes can however be easily included by computing the meridional circulation on moist isentropic surfaces. In midlatitudes, it generates is a circulation that is twice as vigorous as the circulation on dry isentropic surfaces but, because moist isentropes are rarely stratified in the lower troposphere, the resulting circulation is difficult to interpret physically. In this presentation, we describe a technique that solves this limitation by identifying clearly the role played by meridional water vapor transport in setting the difference between the circulations on moist and dry isentropes.

 

The technique we present relies on a careful decomposition of the isentropic mass flux joint distribution. The mass flux joint distribution records mass fluxes according to their potential temperature and equivalent potential temperature. It was computed for the ERA40 reanalysis dataset using

where represents a zonal and temporal mean, 𝛿(·) corresponds to the Dirac delta function, is the surface pressure, 𝑣 is the meridional velocity, 𝜃 is the potential temperature and is the equivalent potential temperature. The meridional coordinate is denoted by 𝜙, the pressure coordinate by 𝑝, the earth's radius by 𝑎 the gravitational acceleration by 𝑔. The projection variables 𝜃′ and correspond to the potential temperature and equivalent potential temperature in the joint isentropic phase-space. The mass flux joint distribution 𝑀 is shown at a fixed latitude in Figure 1. It lies entirely below the main diagonal since the difference is proportional to the moist air specific humidity. The poleward --- or positive --- mass fluxes are found at 315K for a wide range of 𝜃, describing moist mass fluxes across the mid troposphere. The equatorward - or negative - mass fluxes, on the other hand, lie on the main diagonal when 𝜃 is above the mean surface potential temperature but lie on the 80% relative humidity curve for . They are therefore dry in the mid troposphere but moist in the lower troposphere.

 

The novel decomposition that will be described in this presentation splits the joint distribution into directional components according to whether mass fluxes are poleward or equatorward. These are obtained by projecting the directional components onto dry isentropes and thus yield a decomposition of the dry isentropic meridional circulation. The resulting projected directional components differentiate between moist poleward flows and dry equatorward flows. In midlatitudes, they uncover fluxes that would otherwise be cancelled in a dry isentropic average. In particular, we observe that the Ferrel cell represented by directional components reaches all the way to the tropopause and deep within the subtropics.

 

Simultaneously, our method produces profiles of equivalent potential temperature corresponding to each one of the directional components. The recovered directional equivalent potential temperature profiles show midlatitudes eddies that have a very asymmetrical distribution of moisture. Poleward components are shown to be uniformly moist and equatorward components are shown to be dry in the mid troposphere but moist in the lower troposphere. The difference in moisture content between the two directional profiles is large and is on the order of two to three times the equivalent potential temperature standard deviation. This indicates that the midlatitudes moist stability can be explained entirely with the poleward profiles of equivalent potential temperature and we show that these profiles are everywhere less stable than profiles obtained from simpler climatological averages. Because these profiles provide more sensitive measures of moist stability, it helps to determine whether deep convective adjustments or slantwise convective motions are more likely in different regions. The technique we present can be used to explain the connection between surface temperature statistics and midlatitudes moist stability.

 

Figure 1: Mass flux joint distribution for DJF at  𝜙=315N. The red dashed line indicates the mean surface potential temperature . The white solid line indicates the curve corresponding to 80% surface relative humidity.