Evaluation of budget analyses during MISMO

Budget analyses of heat and moisture using networked sounding data (e.g. Yanai et al. 1973) is a classic tool to estimate the thermodynamic impact of convection on the atmosphere. In the MISMO field experiment (Yoneyama et al. 2008), sounding data from three sites was used to compute the atmospheric budgets (Katsumata et al. 2009). The present study evaluates the validity of these budget analyses by comparing budget-derived rainfall to independent estimates from satellites and in-situ rain gauges. In the period leading up to the convectively active phase, the comparison reveals that the budget estimates agree well to those from the other sources. However in the convectively active period, large discrepancies are found among the estimates on the scale of a few to several days. To understand this behavior, the budget analysis is simulated using the flow field generated by a linear model response to MJO-like heating. The results suggest that the discrepancy in rainfall estimates is related to erroneous divergence due to the inability of the triangle configuration of the MISMO sounding network to properly capture the divergence by the rotational component of the wind. Further analyses using the model fields with four sounding sites in a square shaped network does not produce the false divergence. Simulation of the budget analyses using output from NICAM, a high-resolution GCM (Satoh et al. 2008), also demonstrates that the re-constructed Q1 from the four sounding network shows smaller errors than with three sounding sites. These results suggest that future field experiments in the tropics should use four or more sounding sites in a square or diamond shaped configuration to better estimate Q1 and Q2, especially to capture the convection/circulation associated with the Rossby-wave component.