The challenge
Following the events of Cyclone Gabrielle in 2023, several Council’s within New Zealand saw a need to move towards improved flood intelligence. Gisborne District Council began a process to develop new hydrology models for several catchments which could sit within a flood forecasting system.
Our work
This Flood Early Warning System was developed for Te Tairāwhiti, Gisborne District, in New Zealand. The forecasting system uses an operational Delft-FEWS platform incorporating six unique catchments configured within URBS hydrology models. It accesses the local and neighbouring Council gauge networks for rainfall and water levels, which are integrated with official MetService published forecast rainfall, and MetOcean tide and storm surge forecasts to predict flooding. The calibration and validation processes looked at optimising rainfall distribution methods (number of influencing gauges), several rating curves derived following significant events, and optimising model start-times for flood forecasting purposes. In operation, Delft-FEWS is used as a data management system for real-time forecasting, with the ability to apply gauge modifications for erroneous or malfunctioning gauges.
