Spatio-temporal variability of precipitation, temperature and agricultural drought indices in Central Italy.

The agricultural sector is probably the one that will suffer most directly from the climatic variations expected at the global level. In particular, the analysis of the changes expected in water availability and demand is fundamental in order to correctly establish both the present water resource management and the definition of new strategies. In this paper the time series of some climatic and agro-climatic indices in the Region of Umbria (Central Italy) have been analyzed with the aim of finding signs of climate changes and identifying the potential impacts on the agricultural water balance. The aforesaid indices include the precipitation, the mean maximum and minimum temperatures (Tmin, Tmax), the mean temperature range (Δ T), the reference evapotranspiration (ET0) and two drought indices, Standardized Precipitation Index (SPI) and Standardized Deficit Index (SDI, based on the difference between ET0 and precipitation). These indices were analyzed with reference to different periods (in particular the average growing and irrigation seasons). Furthermore, more specific information was obtained by analyzing the simulated water requirement (CWR) and evapotranspiration deficit (ED) of two typical annual crops (corn and sunflower). The time series of the indices were quantified for 38 stations in the region and they were then analyzed with non-parametric tests both at single sites and at the regional level. The tendencies in cumulated precipitation are generally decreasing (particularly during the wet period) and they are also characterized by a defined spatial pattern. The rainfall reduction during the irrigation season, although less widespread, could have the most important practical consequences. The significant trends detected for both the Tmax and Tmin are mainly positive, and they are more evident for Tmin, often resulting in a reduction of Δ T. ET0 shows a prevailing stationary condition due to the counteracting effects of the prevalent reduction observed for the Δ T and the increment of the mean temperature. At any rate, with reference to the irrigation season, ET0 trends are mainly positive. The results obtained for SPI and SDI are in accordance with the tendencies of non-standardized indices with an expectation of extreme drought event occurrence doubled or even tripled over a 30-year time span. Finally, the analysis of the CWR for corn and the ED for sunflower shows a relevant presence of significant positive trends whose impacts can be estimated in respective mean increments of about 23% and 44% over a 50-year time span.