Long-Period Tidal Forcing of Indian Monsoon Rainfall: An Hypothesis

Abstract

This paper examines the nature of the periodic components observed in the interannual variability of June rainfall in northern India by using an eigenvector analysis of the spectra of the June rainfall record (1895?1975) and an eigenvector analysis of the precipitation data itself for stations in that region. The first eigenvectors of these analyses have similar spatial and spectral characteristics which indicate that in June the atmosphere in northern India responds strongly at two frequencies, 0.05 and 0.26 year?1. These two frequencies match the two dominant frequencies in the spectrum of a time series (1895?1975) of the mean monthly soli-lunar tidal potential at the latitude of northern India. It is hypothesized that tidal effects modulate the advance of the monsoon ?front?, producing some of the observed interannual variability. This hypothesis has been tested by using the tidal frequencies to predict June rainfall a year in advance. The success rate of these year-in-advance forecasts in northern India, on independent data, significantly exceeded that expected by chance or predicted by interannual persistence, suggesting that mechanical tidal forcing might be a useful additional long-range forecast tool.