VALUTAZIONI DI CONDIZIONI DI SICCITÀ IN CALABRIA

Observations are essential in climate monitoring since they are the basis for assessing century-scale trends, for the validation of climate models, as well as for detection and attribution of climate change at a regional scale. Precipitation, in particular, is a subject of special concern, being the main component of the global water cycle, a major contributor to extreme events, and a crucial parameter for water resources management. Observation of precipitation is based on ground rain gauges, and on weather radar and satellite retrievals. While rain gauges generally produce the most reliable results in observations, they are often sparsely distributed. Thus, they are not present in adequate numbers to explain precipitation features and provide data to resolve precipitation processes in simulation studies, especially in regions of complex orography and scarce human settlements. This means that rain gauges provide pointwise estimates that may be not fully representative of the area, especially for large areas with few observations. Satellite retrievals and climate reanalysis have thus been used to create regular data grids to fill-in missing observations and to address the scarcity of stations in ungauged regions. A climate reanalysis, combining model results with observations at regular grids, is often produced for all locations on earth, and spans a long time period that can extend decades back. In this paper, drought conditions in Calabria (southern Italy) have been examined through Climate Hazards group InfraRed Precipitation with Station data (CHIRPS), a satellite-based dataset produced at a 0.05° resolution grid (approximately 5 km). The skills of CHIRPS in reproducing correctly the precipitation patterns of the area were validated for the climatological test timeframe 1981-2010 against the registrations of 79 rain gauges of the Multi-Risk Functional Centre of the Regional Agency for Environmental Protection (Regione Calabria). CHIRPS was tested through the Pearson correlation coefficient, and compared to the results of another gridded dataset, European Climate Assessment Dataset's E-OBS reanalysis gridded product, through adimensional metrics based on mean and standard deviation errors. CHIRPS cells showed a correlation of 0.94 with the station data, very close to 1; the result for E-OBS was 0.96. The adimensional metrics showed that CHIRPS proved better than E-OBS in regards to mean and standard deviation errors. The CHIRPS dataset was then used to evaluate meteorological drought values by means of the Standardized Precipitation Index (SPI), evaluated on 3 and 12 months, for the period 1981-2018. The SPI is based on precipitation alone; its computation for any location is based on the long-term precipitation record accumulated over the selected time scale. The long-term record is fitted to a probability distribution, usually a Gamma distribution, which is then transformed into a normal distribution through an equalprobability transformation. It is generally agreed that the SPI on short-term scales describes drought affecting vegetation and agricultural practices, while on long-term scales it is a broad proxy for water resource management. Results showed higher frequencies of severe and extreme drought conditions in the first half of the study period, consistently with results obtained in previous studies by means of rain gauge data. Moreover, the high spatial resolution of the CHIRPS dataset allowed to make considerations for the whole Calabrian territory, highlighting the detection of the highest frequencies of severe and extreme drought conditions, especially in winter, in some specific areas of Calabria, among which those with the highest altitudes. These areas are usually more interested by rainfall events and thus are