The use of historical data to estimate the frequency of rare events: some practical challenges
24th November 2017, 3:30 pm – 4:30 pm
Main Maths Building, SM3
The estimation of the frequency of rare events is a key step in the designing and maintenance of engineering structures and in the development of natural hazards protection schemes. This is typically done using some statistical procedures which use the information on the highest records of some form of systematic measurements. Given the relatively short time span for which reliable measurements on natural processes are available, it is often the case that information of several recording stations are pooled together to give more reliable (i.e. less variable) estimates. Another possible avenue to reduce the uncertainty around estimates of the magnitude of rare events obtained with extreme value statistical models is the use of historical data. This talk will present some statistical models to make use of historical data for a case study in the field of flood frequency estimation, discussing the challenges connected to the use of these models by engineering practitioners. In particular the issue connected to identifying the time coverage of the available historical period is discussed and re-framed as the classical "German tanks" problem.