Evidence for hydrometeor storage and advection effects in DYNAMO budget analysis of the MJO

Large-scale atmospheric budget quantities are computed using an objectively analyzed gridded dataset based primarily on atmospheric soundings collected during the Dynamics of the Madden-Julian Oscillation (DYNAMO) field campaign. Conducted over the central Indian Ocean during the October to December 2011 period, DYNAMO observations captured three robust Madden-Julian Oscillations (MJOs). Using the traditional budget approach of Yanai et al. (1973) with independent estimates of surface evaporation, surface rainfall Po is computed as a residual from the vertically integrated form of the Q2 (apparent moisture sink) equation. This Po field is spatially averaged over the Gan radar domain and compared with rainfall estimates from the SPOL radar operating in Gan for the DYNAMO period. To isolate the MJO signal, these rainfall time series are low-pass (LP) filtered and a three MJO composite is created based on the time (to) of maximum LP-filtered rainfall for each event. Comparison of the LP-filtered rainfall composites shows that budget estimates of rainfall overestimate radar rainfall by ~15% in the MJO build-up stage (i.e., in the days prior to time to) and underestimate radar rainfall by ~8% in the MJO decay stage (i.e., days after time to). In addition, composite analysis of LP-filtered satellite and ground-based estimates of cloud area and liquid and ice water paths (LWP and IWP) show an increase in these fields in the MJO build-up stage and a corresponding reduction in the MJO decay stage. A more precise formulation of the moist thermodynamics reveals that the traditional budget method used to compute Po neglects the effects of hydrometeor storage and advection in computing rainfall. Using satellite and ground-based observations, this presentation will attempt to estimate the magnitude of these neglected effects and determine to what degree they contribute to the rainfall differences noted above.