An Analogy toward the Real-Time Multivariate MJO Index to Improve the Estimation of the Impacts of the MJO on the Precipitation Variability over Iran in the Boreal Cold Months

Abstract

In addition to the tropical data, the OLR/U850/U200 data on a restricted region in southwestern Asia (17.5°N-35°N, 30°E-60°E) has been incorporated into the calculation of the Real-time Multivariate MJO (RMM) index. Introducing this data modifies the leading EOFs of the RMM index slightly and rotates them 14° eastward as a consequence of the change in the column variances of the RMM anomaly matrix at 30°E-60°E longitudes. However, the leading principal components of the new index, which is called Regionally Modified RMM (hereafter, RM-RMM), remain approximately unchanged. Owing to this close relationship between the two indices, the 14° phase-shift is considered as a basis for comparing the influences of the intraseasonal tropics-southwestern Asia variabilities on the precipitation over Iran with those of the MJO. Accordingly, by generating the spatial patterns of the precipitation anomaly, precipitation occurrence probability and other fields based on every active phase of the RM-RMM and RMM indices, the significant differences are revealed between the corresponding phases of the two indices for the phases 1, 3, 5, 7, and 8, separately. These results ameliorate the estimation of the impacts of the MJO on the precipitation and indicate that the RM-RMM index could be used as a companion to the RMM to forecast the precipitation in Iran. Remarkably, the precipitation forecast skill of the lag-regression model based on the RM-RMM index exceeds that of the model based on the RMM for certain time lags during the phases 1, 3, 5, 7 and 8 of the two indices.