Description
Available from Delft-FEWS version 2023.02.
Computes from an ensemble the probability of occurence of an event.
Events are defined as
- the exceedance of a value (e.g., "water level higher than 3.0m") > a
- window between two values (e.g., "water level between 2.0m and 3.0m") > a && < b
- non-exceedance of a value (e.g., "discharge less than 800 m3/s") < b
eventprobability range can consist of a lower and upper boundary (both are optional) and can be configured in 3 ways:
Boundary values directly as a value, an attribute or obtained from a timeseries.
<event><lower><value>10</value></lower></event>
<event><lower><value>@my_lower_bound@</value></lower></event> ,properties are also possible using $my_lower_bound$ instead
<event><lower><variableId>my_lower_bound_timeseries</variableId></lower></event>
Use <inclusive> true for strict lower and upper bounds (>= and <=)
Value:
<transformation id="compute_probability_of_exceedance_of_spi_equal_to_two"> <statisticsEnsemble> <eventProbability> <inputVariable> <variableId>spi</variableId> </inputVariable> <includeMissingValues>false</includeMissingValues> <event> <lower><value>2</value></lower> <upper><value>5</value></upper> </event> <outputVariable> <variableId>probability_spi_gte_2</variableId> </outputVariable> </eventProbability> </statisticsEnsemble> </transformation>
Attribute:
<transformation id="compute_probability_of_exceedance_of_spi_equal_to_two"> <statisticsEnsemble> <eventProbability> <inputVariable> <variableId>spi</variableId> </inputVariable> <includeMissingValues>false</includeMissingValues> <event> <lower><value>@lowerEvent@</value></lower> <upper><value>@upperEvent@</value></upper> </event> <outputVariable> <variableId>probability_spi_gte_2</variableId> </outputVariable> </eventProbability> </statisticsEnsemble> </transformation>
Timeseries:
<transformation id="compute_probability_of_exceedance_of_spi_equal_to_two"> <statisticsEnsemble> <eventProbability> <inputVariable> <variableId>spi</variableId> </inputVariable> <includeMissingValues>false</includeMissingValues> <event> <lower><variableId>lower_boundaries</variableId></lower> <upper><variableId>upper_boundaries</variableId></upper> </event> <outputVariable> <variableId>probability_spi_gte_2</variableId> </outputVariable> </eventProbability> </statisticsEnsemble> </transformation>
Example
In below example, the probability of three events is computed.
- Event 1 is the event where Q.forecast < 10 m3/s
- Event 2 is the event where 10 <= Q.forecast < 20 m3/s
- Event 3 is the event where Q >= 20 m3/s.
In this example, the events are chosen in such a way that they are mutually exclusive and collectively exhaustive. The combined probability should therefore sum to 1. The configuration file for this example can be downloaded from here. A timeseries plot for the results is shown below.