Hale Cycle and Indian Drought and Flood Area Indexes

Abstract

The Indian drought and flood area indexes (DAI and FAI, respectively) have been suggested to be related to the Hale cycle of sunspot numbers by Bhalme and Mooley (1981). They studied the dependence between sunspot numbers, DAI, and FAI series by using spectral, cross‐spectral and dial analyses. The inference problems and the lack of dependability of results associated with crossspectral analysis of autocorrelated time series are well known. Consequently, in this study, the relationships between sunspot number, FAI, and DAI are studied by using more accurate methods of investigation of causal connections between time series and the results are reported. Univariate ARMA models are fitted to sunspot number, double sunspot cycle, DAI, and FAI series and these models are validated. The cross‐correlation properties of residuals from these models are studied. Statistical tests developed by McLeod (1979), Pierce (1977), and Haugh (1976) are used to investigate the causal connections between sunspot numbers and DAI and FAI series. The results indicate that there is no significant relationship between sunspot number and FAI series. There is a consistent, statistically significant relationship between sunspot number and mild and moderate DAI series. However, these relationships are not strong enough to be used in forecasting droughts.