Integration of stochastic and statistical approaches for drought risk estimation through long-length timeseries

Droughts are among the most intense and impactful natural hazard-related disasters facing humanity. The European Commission’s adaptation strategy report (2021) highlights that water scarcity increasingly disrupts a wide spectrum of socioeconomic aspects, ranging from agriculture and food industry to enhancing social and gender inequalities and consequently leading to human, material and economic losses. Therefore, efforts for creating resilient societies, able to deal with climate induced hazards, are high on the agenda both in European and global level. For achieving this universal objective, one of the primary steps suggests deeper understanding of the natural processes responsible for the hazards. In the case of droughts, the detailed study of precipitation patterns and the use of appropriate stochastic simulation methods, which capture the inherent characteristics of natural processes, are crucial for drought risk estimation and forecasting. In this work, we use the state-of-the-art stochastic modeling framework CoSMoS, which implicitly, in terms of the autocorrelation function, and explicitly, in terms of the probability distribution function, adequately simulates the expected variability and interdependence characterizing precipitation records. The stochastic scheme is applied to the precipitation time series of Bologna, which; being one of the longest rainfall time series worldwide, provides a significant advantage in the field of stochastic generation. Following a Monte Carlo simulation approach, 500 synthetic precipitation time series of 100 years each are generated and subsequently analyzed applying run theory to estimate drought risk, frequency and duration, across the city of Bologna and the adjacent provinces. Rather than relying on urgent adaptation measures during a water crisis, the findings and generally the application of the methodology followed in this study, foster a proactive approach in drought management and offer valuable insights in urban and water resources planning, public awareness initiatives, insurance risk assessment and encourage legislative amendments. By integrating probabilistic and statistical methods in drought risk analysis, this work contributes to the global demand to strengthen the resilience of societies against climate related risks.