disaster risk

disaster risk = hazard + vulnerability + exposure
exposure = people, property, system or other elements present in hazard zones that may be affected in potencional losses

3 modules

looks at the physical characteristics of potential disasters
estimating potential future losses - need to create a catalogue of possible future events (possible f.e. hurricanes, that may strike, their intensity, the likelihood they will strike)
for those catalogues are simulated statistical and physical models
collection of samples for potential realisations what could happen are called ensembles
thus any year in the catalogue gives a snapshot of the potential hurricane experience in a single calendar year

accesses to damageability of buildings and their contents
after simulating an event of a given magnitude, the damage it does to buildings must now be computed
need to have indicators, what building will be damaged during hurricane (= certain building characteristic) → will get vulnerability and damage ratio
- damage ratio = how much money will I have to pay for repair
small differences in building characteristics and site-specific effect (wind in this case) can cause different damage effects
- → to solve this, there is no particular value for damage ratio, but the whole distribution of possible values
- after i will do the mean distribution of these values = mean damage ratio
- graph = function of intensity (wind for hurricanes)

damage ratio for specific event s multiplied by the building replacement value to obtain the loss distribution
this module computes the combined loss distribution of all buildings through a process known as convolution
2 loss distributions for 2 locations → process is completed for all other possible combinations and thus the convolved loss distribution for the total loss of the two locations is obtained

EP Curve
describes the probability that various levels of loss will be exceeded
an EP curve is generated by running the catalogue against exposure (buildings) and obtaining losses for each event and year
The events are then grouped by year (the reader should recall at this point that each simulated event has a simulated year to which it is associated) to determine the loss-causing events for each year. The total mean loss for each year is then found by adding the mean losses for each event together. It should also be noted that the mean of the convolved loss distribution for a year is the same as the sum of the mean losses of each event in that year. The losses are then sorted in descending order and plotted to give the exceedance probability and corresponding loss at that probability. The EP curve is the basis upon which insurers estimate their likelihood of experiencing various levels of loss
possibility to invert this curve → will get return period
exceedance probability 1 % == “1 in 100 years storm”

One way for insurers to assess the potential payouts that could be required at various return periods is through plotting percentiles around the EP curve