@@ -14,6 +14,6 @@ In that workshop, we fitted a right-tailed Gumbel distribution to observations o
...
@@ -14,6 +14,6 @@ In that workshop, we fitted a right-tailed Gumbel distribution to observations o
**But what happens if my main interest is the tail?**
**But what happens if my main interest is the tail?**
In Engineering and Geosciences, typically there is an interest in the tails of the distributions. For instance, flood protection systems will be designed to withstand extreme rainfall events or extreme discharges, not only daily conditions. These extreme events are located in the tails of the distribution. Moreover, extreme events are typically scarce in our datasets. The available timeseries are usually short (e.g.: 20 years) in comparison with the design events that the system needs to withstand (e.g.: 1,000 years event). **Extreme Value Analysis (EVA) focuses on those events located at the tails of the distribution and provides a framework to identify and model the stochastic behavior of extreme events so events which have not been observed can be inferred.**
In Engineering and Geosciences, typically there is an interest in the tails of the distributions. For instance, flood protection systems will be designed to withstand extreme rainfall events or extreme discharges (low exceedance probabilities), not only daily conditions (high exceedance probabilities). These extreme events are located in the tails of the distribution. Moreover, extreme events are typically scarce in our datasets. The available timeseries are usually short (e.g.: 20 years) in comparison with the design events that the system needs to withstand (e.g.: 1,000 years event). **Extreme Value Analysis (EVA) focuses on those events located at the tails of the distribution and provides a framework to identify and model the stochastic behavior of extreme events so events which have not been observed can be inferred.**
In the following chapters, the formal concept of extreme will be introduced, as well as the techniques to sample them without a dataset and probabilistically model them. At the end of these chapters, you can also find some supplementary videos, in case you prefer that format.
In the following chapters, the formal concept of extreme will be introduced, as well as the techniques to sample them without a dataset and probabilistically model them. At the end of these chapters, you can also find some supplementary videos, in case you prefer that format.
