Abstract In this study, the annual and seasonal characteristics of meteorological drought are investigated using complex network theory. The analysis is based on the initial reconnaissance drought index (RDI), which is defined as the ratio of precipitation to potential evapotranspiration. This time series is subsequently transformed into a complex network through the visibility graph algorithm. The proposed framework is evaluated against the widely used standardized RDI, which assumes an underlying probability distribution based on the estimated mean and standard deviation. By examining the relationship between the standardized RDI and the topological properties of the resulting network, it is demonstrated that closeness centrality can serve as an effective indicator of drought intensity, while betweenness centrality provides useful information for detecting drought cyclicity and temporal transitions. A notable advantage of the proposed methodology is its robustness to non-stationary time series, whereas the standardized RDI is sensitive to deviations from stationarity.
Vangelis et al. (Fri,) studied this question.