Moist static energy budget during CINDY/DYNAMO

We present an analysis of the moist static energy budget over the sounding array deployed in the equatorial Indian ocean during the CINDY/DYNAMO field program. Data from the sounding array are used to compute advection and tendency terms; OAFlux and CERES data sets are used for surface turbulent fluxes and radiative fluxes respectively. The same analysis is also performed in parallel using the ERA Interim Reanalysis data set. The budgets are complemented by some analysis of the large-scale flow structures responsible for the advection. Three active phase MJO events are studied: one in late October, one in late November, and one in late December; sounding data are available for the first two.

The results show that column radiative heating maximizes in phase with MSE anomalies, acting to reinforce them. Surface turbulent enthalpy fluxes do as well, though with some lag on average (as expected). The magnitude of turbulent flux variations is more variable from event to event than is the magnitude of radiative flux variations, however, and is smaller than radiative flux variations in two out of the three events.

In all events, horizontal advection is relatively moistening in the build-up to the active phase, drying in the later part of the active phase itself, although the absolute values of the horizontal advection are different from event to event. Large moistening by horizontal advection occurs in the build-up to the October event, while large drying by horizontal advection occurs during the rapid decay of the November event. Both of these are consistent with a significant role for horizontal advection in the eastward propagation of the MJO, but in different ways.

The horizontal advection variations result largely from variations in zonal advection of moisture in the 600-900 hPa layer, although the ultimate source of the dry air appears (by inspection of maps) to be subtropical.

The relative magnitudes of horizontal and vertical advection also vary across the MJO events, but the picture that emerges overall is of a significant role for both in MJO propagation. Analysis of the normalized gross moist stability clearly shows an increase in vertical advection from suppressed to peak active phase. This appears to result from a deepening layer of ascent as the active phase matures.