Strong and weak wet monsoon seasons are defined based on observed variability in All-India Rainfall. During ENSO years, differences in the hydrologic cycle between strong and weak seasons are characterized by a substantial increase in easterly moisture transport across the western Pacific Ocean associated with trade winds which are both strengthened and moistened. A large north-south dipole in P-E differences is present over the western Pacific Ocean with negative values lying north of the equator. PW differences exceeding 6 mm are evident over the northern Arabian Sea and Indonesian archipeligo. Corresponding SST differences are negative over the Arabian Sea and far southen Indian Ocean, and are positive over the western Indian Ocean as expected from ENSO's cold phase.
By isolating the differences between strong and weak monsoon years in which SST anomalies in the Nino-3 region are small, a distribution of anomalies is identified which is preseumed independent of ENSO forcing, a presumption supported by the strikingly distinct nature of the anomalies from the aforementioned ENSO related patterns. The distribution of anomalies is characterized by strong moisture transport in and around the monsoon gyre with cyclonic, though largely non-divergent, transports in the northwestern subtropical Pacific Ocean. Transport anomalies in the western Pacific basin are westerly, in stark contrast to the aforementioned ENSO related anomalies. P-E anomalies are significant over the African continent and Indian Ocean and over both domains constitute zonally oriented dipoles. Evaporation anomalies over the Indian Ocean are positive and widespread. Precipitable water anomalies are strongest over northeastern Africa, where they reach 12 mm, or approximately 30% of the mean, and are modest (<3 mm) over the Arabian Sea and other ocean areas. SST anomalies show warm ocean temperatures throughout much of the southern Indian Ocean while cool anomalies are present in the Atlantic Ocean basin and western northern subtropical Pacific.
The identified anomalies in the hydrologic cycle suggest that though ENSO phase is importantly associated with variations in the monsoon hydrologic cycle, its influence does not explain nor resemble anomalies during non-ENSO years. Rather, a distribution of anomalies markedly distinct from those during ENSO years is identified in the Indian and Atlantic Ocean basins and over the African continent. An understanding of the causal relationships which link the identified regions may be gained from further model studies and holds the promise of improved interannual monsoon prediction.3>