Pre-Burst Excess

[hayden...@aecom.com]

Hello,

How is the pre-burst excess depth applied to the burst when the median pre-burst depth exceeds initial loss?

Is it applied entirely to the first time-step, or distributed?

A catchment I am working on in the Wet Tropics Region has an initial loss of 67.0mm and a median pre-burst depth of 119.5mm for the 1% AEP, 12hr event.

Applying the excess 52.5mm to the first 30min time-step would not be appropriate.

It has been mentioned that a more appropriate application of excess pre-burst is to match the intensity of the first design storm time-step for X amount of time prior to the beginning of the design storm, so that the total rainfall over the period X is equal to the excess pre-burst depth. 

Is this a feature Storm Injector will be able to support?

Regards,

Hayden

[Chris Ryan]

Hi Hayden,

Yes, currently pre-burst excess is added to the first timestep so it may produce exaggerated results if the excess is large as in your case. These are not solutions but you could try:

• manually editing the amount of pre-burst excess in the Storms grid by changing the size of the negative number in the IL (mm) column
• changing the pre-burst setting to burst loss = 0, entering a Prefix (such as 'no_excess' clicking Create Storms >> Add to existing. This will allow you to see how the results differ with and without pre-burst excess applied. I did some testing in high pre-burst excess areas and found that it didn't affect the peak results that much since the pre-burst excess was still low compared to main burst. However, volume was definitely increased. It would also depend on the catchment size and critical duration.
• There is also some debate about applying the ARF to pre-burst depths and the latest version of Storm Injector allows this to be done as well.
  

Pre-burst temporal patterns is definitely something that I hope ARR will provide some advice on sooner rather than later...

Regards,

Chris

[Nic Smith]

Hi Chris

Further to pre-burst excess topic below, is it practical to include an option in SI (similar to XPRAFTS) that applies the pre-burst rainfall over a specified number of timesteps prior to the burst to avoid the excess pre-burst (after IL is removed) 'spiking' flows in the early part of the hydrograph? Not such an issue for shorter durations (where IL dominated over pre-burst) but in one case I currently have, the 6 hour events for a particular catchment are becoming dominant due to the large pre-burst excess being applied to the first time step. The 6 hour duration pre-burst is close to 110mm (less 42mm IL) leaving 68mm added on to the first timestep, causing a significant . (South East QLD). Removing the excess (via setting burst loss =zero) removes the influence of excess pre-burst altogether which skews the bias back towards the burst only (this varies across the AEP and duration range given the variability within the mean pre-burst depths).

Also, does the embedded burst smoothing tool reduce these spikes or does it only smooth the temporal patterns prior to pre-burst excess being added?

Thanks

Nic 

On Wednesday, November 7, 2018 at 3:28:45 PM UTC+10, Chris Ryan wrote:

Hi Hayden,

Yes, currently pre-burst excess is added to the first timestep so it may produce exaggerated results if the excess is large as in your case. These are not solutions but you could try:

• manually editing the amount of pre-burst excess in the Storms grid by changing the size of the negative number in the IL (mm) column
• changing the pre-burst setting to burst loss = 0, entering a Prefix (such as 'no_excess' clicking Create Storms >> Add to existing. This will allow you to see how the results differ with and without pre-burst excess applied. I did some testing in high pre-burst excess areas and found that it didn't affect the peak results that much since the pre-burst excess was still low compared to main burst. However, volume was definitely increased. It would also depend on the catchment size and critical duration.
• There is also some debate about applying the ARF to pre-burst depths and the latest version of Storm Injector allows this to be done as well.
  

Pre-burst temporal patterns is definitely something that I hope ARR will provide some advice on sooner rather than later...

Regards,

Chris

[Chris Ryan]

Hi Nic,

The option to apply pre-burst excess over a number of timesteps is something we are considering for future releases. We are a little hesitant in fear of over complicating things. In answer to your other questions, embedded burst smoothing does not consider pre-burst excess. So, I guess it's possible that an embedded burst may be created by the addition of the pre-burst excess if using this option.

Regards,

Chris