Estimation of future Depth-Duration-Frequency relationships from hourly climate projections In Sicily (Italy)

Código

I-EBHE0005

Autores

Gaetano Buonacera, Nunziarita Palazzolo, Antonino Cancelliere, David Johnny Peres

Tema

WG 1.10: Hydrologic Design - Solutions & Communication

Resumo

Rainfall Depth-Duration-Frequency curves (DDFs) are an essential tool for engineering hydrology, as they are widely needed for the risk based design of flood defense infrastructures and other purposes. Climate change is claimed as a possible factor for changing extreme precipitation, so there is a need of methodologies for the derivation of DDF curve considering a non-stationary context. Climate model projections are commonly used to understand how climate may change depending on specific green-house emission scenarios defined at the global level. Given the practical relevance of DDFs, there is also a need for simple and operational methodologies that can be applied at the national level. Within this context, in the present work, a methodology for the derivation of future DDFs is introduced. The methodology assumes simple scaling for short duration precipitation and capitalizes on the availability of hourly projections. It involves the application of change factors, derived from climate models, to the averages of annual maxima series (AMS) of precipitation accumulated over various durations (1, 3, 6, 12, and 24 hours) and to the moments of the dimensionless precipitation series, built by dividing the mentioned AMS by the averages of the respective reference duration. In the present work, a set of Regional Climate Models (RCM) from the EURO-CORDEX initiative, characterized by hourly temporal resolution, are used. The direct utilization of hourly projections allows to avoid the application of temporal disaggregation techniques. Change factors are calculated through the analysis of the AMS derived from control and future hourly precipitation from the RCMs. Regarding the future, we consider two distinct emission scenarios of the Fifth Climate Model Intercomparison Project (CMIP5), namely RCP (Representative Concentration Pathways) 4.5 and 8.5, for three periods, namely: near future (2021-2050), middle future (2051-2070), and far future (2071-2100). An application to rain gauges in Sicily (Italy) is then carried out. Results of the application reveal an upward trend in future DDFs, particularly pronounced in the RCP 4.5 scenario and during the far future period. This behavior can be attributed to a predicted intensification in the variability of rainfall events. Depending on the specific geographic location, chosen emission scenario, and future time period, future Depth-Duration-Frequency (DDF) curves exhibit a significant degree of variability inherited by the uncertainty related to the projections. In some cases, projected return periods are more than double than those observed.