Richard M. Gorman

The significant wave height (H_s) variability caused by wind anomalies associated with the co‐occurrence of the Madden‐Julian Oscillation (MJO) and El Niño–Southern Oscillation (ENSO) was investigated in the New Zealand region. For this purpose,H_sand wind anomalies composites were created using 23 years (1979–2002) of modelled data during November–March periods, when simultaneous ENSO‐MJO phase pairs are potentially most active. The results show striking features: El Niño‐related wave conditions (which consist of increasedH_salong the west and south coasts of New Zealand) are reinforced during MJO phase 8, whereas the wave conditions associated with La Niña (which consist of largerH_salong the north coast) are enhanced during MJO phase 6; Similar wave anomalies are generated during opposing ENSO phases (La Niña and El Niño) when these are combined with MJO phases 3 and 5, respectively; The majority of statistically significantH_sanomalies disappear from the study area during El Niño‐MJO phase 6, El Niño‐MJO phase 2, and La Niña‐MJO phase 4, showing the neutralizing nature of some phase combinations; Finally, negativeH_sanomalies are experienced off the country's west coast during El Niño‐MJO phase 4, in contrast to the positive anomalies expected during El Niño events. These results clearly show the importance of remote forcing to wave anomalies in the New Zealand region, and highlight the need to assess atmospheric and oceanic conditions considering multiple climate oscillations.

Most geophysical models include many parameters that are not fully determined by theory, and can be “tuned” to improve the model's agreement with available data. We might attempt to automate this tuning process in an objective way by employing an optimisation algorithm to find the set of parameters that minimises a cost function derived from comparing model outputs with measurements. A number of algorithms are available for solving optimisation problems, in various programming languages, but interfacing such software to a complex geophysical model simulation presents certain challenges. To tackle this problem, we have developed an optimisation suite (“Cyclops”) based on the Cylc workflow engine that implements a wide selection of optimisation algorithms from the NLopt Python toolbox (Johnson, 2014). The Cyclops optimisation suite can be used to calibrate any modelling system that has itself been implemented as a (separate) Cylc model suite, provided it includes computation and output of the desired scalar cost function. A growing number of institutions are using Cylc to orchestrate complex distributed suites of interdependent cycling tasks within their operational forecast systems, and in such cases application of the optimisation suite is particularly straightforward. As a test case, we applied the Cyclops to calibrate a global implementation of the WAVEWATCH III (v4.18) third-generation spectral wave model, forced by ERA-Interim input fields. This was calibrated over a 1-year period (1997), before applying the calibrated model to a full (1979–2016) wave hindcast. The chosen error metric was the spatial average of the root mean square error of hindcast significant wave height compared with collocated altimeter records. We describe the results of a calibration in which up to 19 parameters were optimised.

Clusters of storm waves contribute disproportionately to coastal erosion hazards because the coastline has insufficient time to recover between events. Here, the change in occurrence of clustered storms and its association with atmospheric oscillation modes were investigated in New Zealand waters using 44 years (1958–2001) of wave hindcast data. First, long‐term averages of cluster parameters (number of storms within the cluster, potential for coastal erosion, and cluster duration) were assessed. Then, the relationships between clustering and the El Niño‐Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Zonal Wavenumber‐3 Pattern (ZW3), Pacific Decadal Oscillation (PDO), and Southern Annular Mode (SAM) were explored through correlation analysis over several timescales. Clusters were more frequently observed to the northeast of New Zealand and on the central eastern coast of the South Island. The most vulnerable regions to cluster‐induced coastal erosion were southern New Zealand and the northwestern coast, which resulted from steady southwesterly swells, although clusters with the longest duration occurred on the east coast of the South Island. Trends suggest that clusters have incorporated more storms, have become more hazardous, and have increased in duration, particularly along the South Island coastline. Although these trends may be sensitive to the reanalysed wind fields used to force the wave hindcast, they reflect trends in the ENSO, PDO, and SAM. Stronger southwesterly winds during El Niño (negative ENSO) and El Niño‐like conditions (positive IOD/PDO) generated more clustered storms mainly on the southwestern coast of New Zealand, whereas increases in clustering were observed on the north coast during La Niña and La Niña‐like conditions (stronger northeasterly winds). Higher occurrence of clustering was also evident on the west coast during the strong atmospheric zonal flow associated with negative ZW3. Lastly, strengthened westerlies related to positive SAM led to increased clustering primarily to the south of New Zealand.

Most geophysical models include a number of parameters that are not fully determined by theory, and can be ‘tuned’ to improve the model's agreement with available data. We might attempt to automate this tuning process in an objective way by employing an optimisation algorithm to find the set of parameters that minimises a cost function derived from comparing model outputs with measurements. A number of algorithms are available for solving optimisation problems, in various software programming languages, but interfacing such software to a complex geophysical model simulation, presents certain challenges. To tackle this problem, we have developed an optimisation suite ("Cyclops") based on the Cylc workflow engine (http://cylc.github.io/cylc/ and https://zenodo.org/badge/latestdoi/1836229) that implements a wide selection of optimisation algorithms from the NLopt python toolbox (Johnson, 2014). The Cyclops optimisation suite can be used to calibrate any modelling system that has itself been implemented as a (separate) Cylc model suite, provided it includes computation and output of the desired scalar cost function. A growing number of institutions are using Cylc to orchestrate complex distributed suites of interdependent cycling tasks within their operational forecast systems, and in such cases application of the optimisation suite is particularly straightforward. As a test case, we applied the Cyclops to calibrate a global implementation of the Wavewatch III™ (v4.18) third generation spectral wave model, forced by ERA-Interim input fields. This was calibrated over a one-year period (1997), before applying the calibrated model to a full (1979–2016) wave hindcast. The chosen error metric was the spatial average of the root-mean-square error of hindcast significant wave height compared with collocated altimeter records. We describe the results of a calibration in which up to 19 parameters were optimised.

This work investigates how the wave climate around New Zealand and the southwest Pacific is modulated by the Pacific Decadal Oscillation (PDO), El Niño‐Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Zonal Wave‐number‐3 Pattern (ZW3), and Southern Annular Mode (SAM) during the period 1958–2001. Their respective climate indices were correlated with modeled mean wave parameters extracted from a 45 year (1957–2002) wave hindcast carried out with the WAVEWATCH III model using the wind and ice fields from the ERA‐40 reanalysis project. The correlation was performed using the Pearson's correlation coefficient and the wavelet spectral analysis. Prior to that, mean annual and interannual variabilities and trends in significant wave height (H_s) were computed over 44 years (1958–2001). In general, higher annual and interannual variabilities were found along the coastline, in regions dominated by local winds. An increasing trend in H_s was found around the country, with values varying between 1 and 6 cm/decade at the shoreline. The greatest H_s trends were identified to the south of 48°S, suggesting a relationship with the positive trend in the SAM. Seasonal to decadal time scales of the SAM strongly influenced wave parameters throughout the period analyzed. In addition, larger waves were observed during extreme ENSO and IOD events at interannual time scale, while they were more evident at seasonal and intraseasonal time scales in the correlations with the ZW3. Negative phases of the ZW3 and ENSO and positive phases of the IOD, PDO, and SAM resulted in larger waves around most parts of New Zealand.

A new marine forecasting system for the New Zealand region has been developed, including global and regional wave models, and regional tide and storm surge models. These form part of an integrated weather-related hazards forecasting capability, which also includes an accurate, data-assimilating high-resolution weather forecasting system and a national-scale flood forecasting system. Predicted variations in sea surface height and depth-averaged current due to tides and storm surge are provided in twice-daily 48-hour regional forecasts. Storm surge is predicted for the New Zealand region using the RiCOM hydrodynamic model on an unstructured grid in depth-averaged mode, using semi-implicit integration in time and a semi-langrangian scheme for advection. The model reproduced the storm surge event resulting from the passage of cyclone “Funa” across New Zealand during January 2008. Root-mean-square (RMS) errors in predicted storm surge height were 4-13 cm. For a second simulated storm, which occurred in Marlborough in late July/early August 2008, the model predicted the storm surge height with RMS errors at Kaikoura and Lyttelton of 7 and 4 cm, respectively. Wave-breaking also contributes to sea levels close to the coast through radiation stresses. The forecast system runs Wavewatch III on nested global and regional structured grids, providing twice-daily 48-hour forecasts. Model developments incorporating wave radiation stresses into the hydrodynamic model forcing are described, and some investigations into the effects of wave forcing on surface setup and coastal inundation are presented. A case study of inundation around the fringes of Hawkes Bay in response to extreme sea levels exacerbated by wave-induced stresses is presented. The case study used the SWAN wave model to provide the near-shore wave characteristics and radiations stresses, and demonstrates the ability of the coupled tide-surge-wave model to simulate and predict coastal flooding on a spatial resolution of a few metres.