Assessing the skill of all-season diverse Madden-Julian oscillation indices for the intraseasonal Amazon precipitation

Madden-Julian Oscillation (MJO) impact on the Amazon intraseasonal precipitation assessed by different MJO indices is investigated through an analysis of composite events and observed case studies. The MJO indices diagnosing skill is described in detail using reanalysis, satellite, and gauge-based gridded rainfall data. Three types of existing MJO indices are considered: (1) OLR-based MJO (OMI index); (2) dynamically- or circulation-based MJO (VPM index); and (3) combined convectively- and dynamically-based MJO (Wheeler-Hendon RMM index). Our results suggest that, in the large-scale (i.e., around the global tropics), even in the regional domain, the MJO OLR-only index well-represent the dynamical and convective features associated with the intraseasonal variability. On the other hand, each index gives diverging results on rainfall characterization over the Amazon Basin (AB). For instance, the cumulative distribution of precipitation for each MJO phase and index depicts considerable differences in the main climatic regions of the AB, indicating a diverging intraseasonal representation for extreme rainfall values. In addition, when event-by-event is assessed, details as well as the identification of events themselves can differ among indices. This characteristic is particularly observed during extreme rainfall events in the AB. A significant percentage of MJO activity is detected only by the MJO OLR-based index. Because the large-scale zonal circulation dominates the dynamically-based indices (RMM and VPM), the MJO impact in these indices is not an appropriate measure of convective MJO activity. Since the convective component of the MJO is our primary objective, the results presented in this study show that the OLR-based MJO index is able to better account for the MJO impacts over the AB. The indices considered in this study are often used for monitoring and forecasting the MJO activity over South America. However, given the dissimilarity of the representation of precipitation in the AB, our findings also support the consideration of a regional index for monitoring and forecasting the MJO impacts.