![Sample our Environment & Agriculture journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5934/TOC-Subject-Sample-2021-Environment-Agriculture.jpg"})
Abstract
Numerical models of the marine wind field and wave generation offer a technique for estimating the severity of extreme wave events. Measured and modelled waves in the Canterbury Bight, New Zealand, were used to examine the hypothesis that extreme conditions can be more confidently estimated by modelling the wave climate over a decade than by analysing shorter measured records. The meteorological conditions causing high waves in the Canterbury Bight were examined by inspecting wind measurements at Christchurch Airport and wave buoy measurements made in 1983–85 near the Rakaia River mouth. Although the arrival of high waves was sometimes associated with wind changes at the airport, peak wave heights generally occurred at least several hours after such wind changes and were associated with south‐westerly or southerly winds in the Bight and a depression south‐east of the South Island. Wave conditions at the buoy site for 1980–89 were modelled using marine winds obtained from a global circulation model, running a wave generation model for the seas around New Zealand, and then applying a refraction model. The decade modelled included two El Niño events and one La Niña event, and the ocean wind climate should therefore have been reasonably representative of the long‐term range of conditions. The model results were verified against the buoy measurements. Estimates of extreme wave heights were then made from 10 years of modelled record and also the measured buoy record. Application of the Gumbel distribution to the monthly maxima gave a significant wave height of 9.5 m for the wave event with an annual exceedance probability of 1 in 50. Model predictions indicate that the most severe waves arrive in the Bight from due south. Estimates of extremes from the measurements and model hindcast were close, indicating that the hindcast represents a population of events similar to those measured.