Extreme Precipitation Analysis
Running a Baseline Scenario
The following is an explanation of the processes involved in conducting an extreme event analysis. Read through the process and look at the sample results and then proceed to conducting your own extreme event analysis.
Click on the icon within the Impact Models tab to start extreme event analysis.
Select a data type to begin:
Then select a site by either toggling in the box or by clicking the site on the associated area map. The selected site will be highlighted in the map.
Associated map. Click on a site to select a station to work with. Note: that the map of associate sites that you will see will be dependent on the site data packages you have loaded in your SimCLIM for Desktop package.
Select a scenario by clicking on the Scenario button.
The Type option provides the flexibility of defining the period when the extreme value is of concern.
The period of interest can be defined by selecting month, season, or by defining a start date and end date as a period, or by fixing a date and with a days range.
Select the months for analysis by clicking on the Month box and making your choice. Select all is a good place to start. Later you can explore seasonal and other results.
By default, the Run for historical period date boxes shows the complete period available. The user can define their period by changing the dates.
Select the extreme event variable by clicking the corresponding tab. If the data is not available, the box in its tab page will be disabled.
The box in the tab page determines how many consecutive days you want to carry out the analysis. The Extreme event definition option will become active if you selected more than one day for your analysis.
Then you will have the choice to select total values (sum of the consecutive day’s data), on the highest value of the consecutive days or on the lowest value of the consecutive days.
You can also select based on the quality of the data you wish to accept in your analysis. It is not uncommon for large data sets to have missing days, months and even years. The amount of missing data can have an impact on GEV analysis.
A simple way to view the data's completeness is to open the Manage Sites tool outside of the extreme event analysis tool as shown below:
Then select the data site you wish to explore for completeness by typing it into the search window. Select the variable you wish to check, in this case precipitation and then click on Quality.
A graph and table will then be displayed showing the completeness of the data file for that particular variable:
The Brisbane Aero site is relatively complete. As can often be the case the first and last year of the data record can be incomplete. With this information you can select the specific dates with full data records when completing your extreme event analysis.
The final option is the PWM calculation order. Theoretically, a high order gives a better model fits towards the extreme end, but at the possible expense of under-performance of the model at the not as extreme part. By clicking run, the model result will show in the table.
The figure shows the extreme daily rainfall event result for Brisbane Airport based on 1959 to 2000 historical observed data. For the selected month (the whole year in this example) with one day rainfall total, the estimated GEV model output is listed in a results table. By clicking on the results table, a graph will be displayed:
Further analysis can be carried out using the design value and risk calculator at the bottom of the graph. In the above example, the result demonstrates that the 100 year-return-period extreme event for 1 day total precipitation is about 315.53 mm. There is a 18% chance during any continuous 20 year period of such an extreme event occurring.
Click on to display more information on return periods and values:
Simply type into the year dialogue box if you wish to see, for example the baseline 200 year and 500 year return period values.
Running a Climate Change Scenario
Return to your dialogue box you prepared for your baseline scenario analysis of extreme precipitation and click on Scenario and the Extreme Event Change Scenario dialogue box will appear:
Select the year and you need not touch the latitude and longitude boxes as they derive this information from the weather file site you have chosen. In this case Brisbane Aero.
You can also select one or more climate model patterns and the RCP of your choice for analysis.
Click and back in the main dialogue box.
Then click on the results table above and a new graph will be displayed that includes the baseline and future scenario outputs:
Click on the Table button to see the numerical display like you did with analysis of the baseline and you can now toggle between baseline and climate change conditions based on the models and scenario you have chosen.
You will see that the 1 in 100 year event for this site under a RCP 8.5 medium sensitivity scenario increased from 315.5to 381.7 mm.