Drought risk for water supply systems based on low-flow regionalisation

This work focuses on low flow indices that are commonly evaluated at gauged sites from observed streamflow time series. Hydrological data are not always available at the site of interest: regional frequency analysis is commonly used for the estimation of flows at sites where little or no data exists. The study is applied to Tuscany rivers discharge dataset, recorded from 1949 to 2008. The area is subdivided into homogeneous regions using an L-moments procedure. The low flow indices are evaluated with deterministic and geostatistical methods. A multivariate model, based on geomorphoclimatic characteristics, is also assessed. For each sub-region a relation connecting low flow indices and geomorphoclimatic characteristics is found. Drought indices show little correlation with water shortage situations that depend also on water storage, demand fluctuation and on the actions carried out to reduce drought effects. For that reason an indicator relating supply and demand is required in order to identify situations of risk of water shortages. An analysis of the relationship between failure of water supply systems and reservoir volumes for the area of Firenze, is performed using Monte Carlo simulations. The reservoir levels and volumes are simulated using time series of the period 1970-2005. Four scenarios (i.e. normal, pre-alert, alert and emergency) associated with different levels of severity of drought are defined. Threshold values are identified considering the probability to assure a given fraction of the demand in a certain time horizon, and are calibrated with an optimization method, which try to minimize the water shortages, especially the heavier. The critical situations are assessed month by month in order to evaluate optimal management rules during the year and avoid conditions of total water shortage.