We have discovered a clear increasing trend in both rainfall and intensity of the hydrological cycle over tropical South America over the past four decades. Three separate data sets -the Global Historical Climatological Network, outgoing longwave radiation, and the NCEP/NCAR reanalysis data- each provide evidence of this increase, giving added confidence to the existence of this inerdecadal increasing trend. Analysis of global circulation patterns reveals that an increase in large-scale convergence over the past 40 years is the root cause of this increased moistening. These results have major implications for future studies of Amazonia:

1) Suppression of impacts of Amazonian deforestation All global and regional-scale numerical studies that overlook influences of interdecadal global convergence change indicate deforestation reduces rainfall and evaporation. Observed trends contrary to these numerical results suggest the interdecadal increase of moisture convergence is of sufficient magnitude to more than compensate for mesoscale drying due to deforestation.

2) Interpretation of impacts of ENSO The interdecadal trend of global-scale divergence enhances (reduces) convergence of water vapor flux toward tropical South America during cold (warm) ENSO phases with concurrent enhancement (reduction) of summer rainfall. The contrast of summer rainfall in the Amazon Basin between the two extreme ENSO phases is not altered, but the interdecadal increasing trend of water vapor flux convergence into this region enhances rainfall during both cold and warm ENSO phases.

Our results suggest that future simulations of the impacts of basin-scale deforestation, the regional hydrological cycle, and the forcing of regional climate of Amazonia must include mechanisms describing both land-surface processes and interdecadal changes of the global divergent circulation.