Cyclone Gabrielle broke all sorts of records, and not in a good way. The National Institute of Water and Atmospheric Research has drilled down into flood flows for more insight into the event. Wairoa Star editor Ann Revington looks into their report . . .
Scientific analysis of Cyclone Gabrielle’s flooding has underlined the extraordinary magnitude of the event.
Modelling of the Wairoa flooding is being peer-reviewed and due to be released shortly.
The National Institute of Water and Atmospheric Research (NIWA) undertook analysis of the flood flows that occurred at 20 river gauge sites across Hawke’s Bay during Cyclone Gabrielle.
It was funded by the Ministry of Business, Innovation and Employment (MBIE).
NIWA’s principal scientist of natural hazards and hydrodynamics, Emily Lane, says the work was to understand Cyclone Gabrielle’s flooding in the context of what was previously known about flood flow values in the region.
“This event reset the scale for flooding in Hawke’s Bay.
“It’s vital we factor this into our understanding of future flood hazards,” said Dr Lane.
“Flood events occur randomly, and so just because a large flood with a high annual recurrence interval has just occurred, there is no guarantee that there will not be another one in the near future.”
The river gauges peak flow rates during the event in the Wairoa catchment, Hawke’s Bay Regional Council has confidence in are Waiau River at Ardkeen (peak flow 1656 m3/s), Waiau at Otoi 838 m3/s, Ruakituri River at Tauwharetoi 998 m3/s.
A rated value was used for Wairoa River at Marumaru (WSP modelling), 4100 m3/s. The rated flow peaked at 4962m3/s, but modelling by WSP suggested an issue with the rating for this level of flow and the lesser value was more appropriate.
Hangaroa at Donneraille Park, 2070 m3/s, was rated from measured flood depth.
“Flood debris was two metres higher than the gauged level. This is the rated value based on that level.”
Pre-Gabrielle, the probability of a flood this size occurring in a given year, called an ARI, was as high as a one-in-1000-year event at some river sites, according to NIWA’s modelling.
An ARI is the average number of years that is predicted will pass before an event of a given magnitude occurs.
For example a 40-year ARI event would on average happen every 40 years.
Post-Gabrielle, that probability has changed to a one in 550-year event – meaning the cyclone has changed the standard going forward.
At 13 of the 20 sites, it was deemed to be the largest flood on record.
‘One piece of the puzzle’
Dr Lane’s summary said the results of the extreme value analysis show an ARI of 400 to 500 years in the Heretaunga Plains region (specifically the Tūtaekuri and Ngaruroro Rivers).
“In many of these locations, prior to Cyclone Gabrielle these events would have been assessed with an ARI of around 1000 years or higher.
“In other locations, such as the Esk Valley, Wairoa and Hangaroa Rivers and Mangaorapa Stream it has a return period of 100-200 years.
“In other locations it was still a significant event but with an ARI of less than 100 years.”
Hawke’s Bay Regional Council chair Hinewai Ormsby welcomed the report and its data as one piece of the puzzle to inform their flood and scheme reviews.
“The modelling reinforces that the cyclone flooding was largely unprecedented.
“The sheer volume of the cyclone’s flooding and its impacts on flood mitigation infrastructure could be compared to having a 100-bed hospital where suddenly 500 patients turn up.
“This report is incredibly valuable in understanding ecosystem dynamics and managing flood risk, in a world that is changing around us.
“It is an essential first step, but not the only step in informing the regional council in its development of future flood resilience designs and mitigations.”
The regional council worked with NIWA and provided regional expertise and understanding of the river’s dynamics, including long-term data and knowledge of historic flood levels.
Chris Dolley, regional council’s group manager-asset management said the NIWA modelling work was extremely beneficial and would allow concept designs for stopbanks and other infrastructure to be fine-tuned.
Danette Olsen, MBIE general manager science system investment and performance, said, “the science sector has expert knowledge to share on the impact of extreme weather events such as Cyclone Gabrielle. The funding provided by MBIE on behalf of the Government will enable us to better prepare for future events.”
Understanding flood risks
An annual exceedance probability (AEP) is the probability a flood of a certain magnitude this year will occur in a given year (i.e. a 100- year ARI flood has a 1 percent AEP).
The occurrence of a rare flood does not reduce the chances of another rare flood within a short time period.
The probability of occurrence is the same every year.
The pre-Gabrielle ARI expresses how likely it was thought this magnitude flood was before it happened (based on available data).
The post-Gabrielle ARI reveals now that we know Cyclone Gabrielle has happened, our best estimate of having a similar-sized flood in the future.
To understand the risk of floods, the regional council uses modelling techniques to understand the probability of a flood event occurring.
The NIWA work provided critical modelling about river flows that occurred during the cyclone so agencies are better able to judge areas at risk from future flooding and better able to predict water depths and velocities should flooding occur.
Understanding the magnitude and frequency of floods is vital in managing flood risk.
At 13 of the 20 measurement sites, Cyclone Gabrielle was deemed the largest flood on record.
The sites where this was not the largest on record were:
■ Ngaruroro River at Whanawhana
■ Tukituki River at Redbridge
■ Waiau River at Ardkeen
■ Waiau River at Otoi
■ Hangaroa River at Doneraille Park
■ Ruakituri River at Tuwharetoa
■ Kopuawhara Stream at Railway Bridge.
The Hawke’s Bay Regional Council will assess the impact for all flood mitigation infrastructure, such as stopbanks. In two areas where the NIWA data is less certain — in Waipawa and Esk Valley — further work is required to understand the flows and implications for flood design mitigations.
It is expected the regional-scale report will inform all current scheme reviews and planning for future schemes, including proposed schemes under the land categorisation.
The report was commissioned by MBIE and prepared by NIWA as a state of the environment document.
The main purpose of the NIWA modelling was to understand Cyclone Gabrielle’s flooding in the context of what was previously known about flood flow values in the region.
The NIWA data is not meant to be definitive, but a useful starting point for agencies to assess the impact on such things as building consents and bridge designs.
• See the report at https://tinyurl.com/3vv284cd
In her report to the Hawke’s Bay Regional Council, February 23, Dr Emily Lane said due to the severity of Cyclone Gabrielle, some river gauges failed, and they did not have a measured peak-flow rate for these locations.
“In these locations, flood models have been developed to replicate the observed flooding and flow estimates from these models have been used in the extreme value analysis.
“Because of the inherent uncertainty in the flow estimates, in some cases more than one value has been used to better understand the uncertainty in the results.
“The systematic record (long-term monitoring information) was used, however, in some locations, there are known large flood events that occurred prior to the systematic record where we have reasonable estimates of the river flow.
“In these locations we have also estimated the flood statistics taking into account those historical events.”
There was also some uncertainty in the maximum flood flow at some sites where there was out-of-bank flow (breaches or overtopping), where water may have been backed up due to debris dams downstream, or where the rating curve has not been adequately assessed for high flows (due to the size of this event compared to previous floods).
For these locations, modelled flood flows were also considered.
Other modelling was undertaken for specific recovery projects on several rivers and estimates of flood flows from this modelling have been used for Tukituki at Red Bridge, Esk River at Waipunga Bridge (PDP 2023), Wairoa River at Marumaru and Waipawa River at RDS.
Modelling of the flooding in Wairoa was performed for regional council by WSP post-Cyclone Gabrielle for the purposes of land categorisation and flood risk management optioneering.
■ The report, “Wairoa River Fluvial Hydraulic Model” (WSP, 2024), is under peer review, but will be released in the near future.
Gisborne Herald readers share their views.