Potential predictability of dry and wet periods: Sicily and Elbe-Basin (Germany)

TLDR: In this article, the authors evaluate a viable tool for the potential predictability of dry and wet spells using principal component analysis (PCA) and find that periodicities ranging from 3.4 to 12 years characterise the SPI signals in both regions and essentially drive the main dry or wet occurrences.

Abstract: The purpose of this paper is to evaluate a viable tool for the potential predictability of dry and wet spells. We select two regions in Europe that have distinct precipitation regimes: Sicily and Elbe basin (Germany). The analysis of dryness and wetness in these regions from 1951 to 2000 is based on the Standardized Precipitation Index (SPI) computed on a long-time scale (two years) and the evaluation of their time-space variability is carried out using Principal Component Analysis. Results suggest that periodicities ranging from 3.4 to 12 years characterise the SPI signals in both regions and essentially drive the main dry and wet occurrences. In Sicily, at variance with the Elbe basin, superimposed to this variability there is also a clearly detectable linear trend that is perhaps related to long-term periodicity. Moreover, the shift in phase found between the common periods implies that often on the longer time scale if the Elbe region has dry conditions, Sicily is wet and viceversa. The reconstruction of the SPI time series by considering the periodicity that greatly contribute to the total power spectrum variance gives good results and provides good opportunities for predictability.