Characterization of Present and Future Droughts in Northern Tuscany (Italy)

Abstract:

Droughts are “creeping” natural hazards that are usually related to extended time periods with rainfall deficiency when compared to average statistics. Characterization of regional, present and future, droughts associated with climate natural variability and/or global changes are important to assess their impact on water resource availability. In this work, we make use of the Standardized Precipitation Index (SPI) to identify and quantify, at different timescales, precipitation deficits and droughts in the northern part of the Tuscany Region (Italy). The study site extends from the Appenine to the Tirrenian Sea and has an area, of about 3000 km². Historical rainfall data collected in the last 90 years are available in gauging stations covering the entire area. We first analyzed the historical drought (in terms of SPI) frequencies in comparison with those of the last decade. It seems likely that an increase in droughts in the last period, at all timescale, can be detected from the available data. Then, to evaluate the effects of natural climate variabilities which, as some Authors as pointed out, can be far greater than climate change signals, we made use of a multi-site weather generator to produce synthetic rainfall time-series that match the statistical properties of the historical data. In this way we provided equally likely alternative realizations of the last decade climate useful to explore the variability and evaluate the uncertainty of the present drought frequencies. Moreover, to characterize the possible upcoming droughts (next 25 and 40 years) we made use of synthetic future rainfall time-series. For this purpose, we modified the weather generator parameters on the basis of available RCM climate model outputs under the RCP4.5 and RCP8.5 scenarios. This study is a first step, for the local water supply company, to understand the need of implementing immediate and/or medium-tem water conservation activities and drought response measures over the area.