Investigating DYNAMO Hypotheses of the MJO Using the YOTC CloudSat-Centric, Collocated A-Train and ECMWF Data Set

The prediction and representation of the Madden-Julian Oscillation (MJO) in numerical models remains inconsistent and uncertain, despite of its importance to tropical weather and climate forecast. Recently the CINDY2011/DYNAMO field experiment has provides valuable in situ observations to improve the understandings and the modeling skills of the MJO. The present study aims to assist in the investigation of the DYNAMO scientific hypotheses regarding to the role of the moist layer and cloud populations for MJO initiation and propagation, from the perspective of a new collocated multi-parameter data set. The data set consists of selected products from the NASA A-Train and ECMWF analyses, all being collocated along the CloudSat footprint globally during 2008-2010. As a contribution to the WMO Year of Tropical Convection (YOTC) research activity, the data set provides observational constraints to the study of clouds and convection. The statistics are analyzed over the CINDY2011/DYNAMO area in the boreal winter MJO season, including the probability distribution function (PDF) and the empirical orthogonal function (EOF) of the vertical profiles of key variables. The characteristics and variabilities with the MJO phase are described for the ambient thermodynamics and dynamics, cloud structure, microphysics, and precipitation. Potential process-oriented diagnostics and constraints are identified to facilitate model evaluation. Sampling biases and uncertainties associated with instrument sensitivity are also discussed. The present study demonstrates the new opportunities provided by the collocated multi-parameter data set, especially those for the improvement and evaluation of the parameterizations associated with the moist convection process and the MJO in the atmospheric models.