Emily Lane
Principal Scientist – Natural Hazards and Hydrodynamics, Earth Sciences NZ (NIWA)
Emily.Lane@niwa.co.nz
Cyclone Gabrielle caused flooding over much of the east coast of the North Island, especially in Hawkes Bay and Tairāwhiti. Understanding the flooding that occurred helps us better prepare for future events. This presentation reports on work undertaken by NIWA (now Earth Sciences NZ) and collaborators in the aftermath of Cyclone Gabrielle to better quantify the flood hazard in Hawke’s Bay and Tairāwhiti.
The 2022-2023 period was characterised by La Niña conditions, which typically bring increased rainfall to New Zealand’s North Island, particularly to the northern and eastern regions. At the time Cyclone Gabrielle approached New Zealand, a significant marine heat wave was present in the waters northeast of the country. Temperature anomaly data from 6 February 2023 showed areas of elevated sea surface temperatures where waters were substantially warmer than typical for that location.
A comprehensive rainfall analysis was developed by synthesising multiple data sources. The analysis combined observations from rainfall gauges operated by Gisborne District Council, Hawke’s Bay Regional Council, and Earth Sciences NZ, along with data from individual private gauges. The three-day rainfall totals showed significant spatial variation, with the most intense precipitation occurring on 13 February 2023.
The flood modelling approach employed a coupled hydrological-hydrodynamic methodology. The analysis focused on specific floodplains within the broader river catchments. Stop bank (levee) breaches represented a crucial component of the flooding process and were incorporated into the model based on field observations and timing estimates. Regional councils expressed significant interest in understanding Cyclone Gabrielle in the context of flood frequency and return periods. Flood frequency analyses were conducted for 20 locations across Hawke’s Bay and Tairāwhiti to characterise the event’s rarity.
Complementary research conducted in collaboration with James Brasington at the University of Canterbury employed LiDAR (Light Detection and Ranging) differencing to quantify sediment erosion and deposition. This technique compared pre-event and post-event high
The comprehensive modelling programme successfully characterised the flooding through integration of observed rainfall data, hydrological modelling, hydrodynamic simulation, and incorporation of stop bank breach dynamics. Model validation using multiple independent datasets, including field measurements, aerial imagery, gauge records, and community-sourced impact information, demonstrated reasonable agreement between simulated and observed flood characteristics.
The event caused substantial geomorphic impacts through bank erosion and sediment deposition, with hundreds of thousands of cubic metres of sediment mobilised along affected river reaches. The role of riparian vegetation in moderating erosion was identified, though with important caveats regarding the potential for large trees to trigger localised bank failures.
This work has enhanced understanding of extreme flood hazards in the region and contributed to improved public information resources through the national flood viewer platform. The findings provide important context for flood risk management, land use planning, and infrastructure design in these communities. Given uncertainties regarding future climate impacts on extra-tropical cyclone frequency and intensity, continued investment in flood hazard research, modelling, and monitoring remains essential for building resilience to similar events that may occur in the future.-resolution topographic surveys to calculate elevation changes across the landscape.