Patrick G. Timko

This paper discusses the assimilation of significant wave height (Hs) observations from the China France Oceanography SATellite (CFOSAT) into the Global Deterministic Wave Prediction System developed by Environment and Climate Change Canada. We focus on the quantification of background errors in an effort to address the conventional, simplified, homogeneous assumptions made in previous studies using Optimal Interpolation (OI) to generate Hs analysis. A map of Best Correlation Length, L, is generated to count for the inhomogeneity in the wave field. This map was calculated from pairs of Hs forecasts of two grid points shifted in space and time from which a look-up table is derived and used to infer the spatial extent of correlations within the wave field. The wave spectra are then updated from Hs analysis using a frequency shift scheme. Results reveal significant spatial variance in the distribution of L, with notably high values located in the eastern tropical Pacific Ocean, a pattern that is expected due to the persistent swells dominating in this region. Experiments are conducted with spatially varying correlation lengths and a set correlation length of eight grid points in the analysis step. Forecasts from these analyses are validated independently with the Global Telecommunications System buoys and the Copernicus Marine Environment Monitoring Service (CMEMS) altimetry wave height observations. It is found that the proposed statistical method generally outperforms the conventional method with lower standard deviation and bias for both Hs and peak period forecasts. The conventional method has more drastic corrections on Hs forecasts, but such corrections are not robust, particularly in regions with relatively short spatial correlation length scales. Based on the analysis of the CMEMS comparison, the globally varying correlation length produces a positive increment of the Hs forecast, which is globally associated with forecast error reduction lasting up to 24 h into the forecast.

Temperature variance and kinetic energy (KE) from two global simulations of the HYbrid Coordinate Ocean Model (HYCOM; 1/12° and 1/25°) and three global simulations of the Massachusetts Institute of Technology general circulation model (MITgcm; 1/12°, 1/24°, and 1/48°), all of which are forced by atmospheric fields and the astronomical tidal potential, are compared with temperature variance and KE from a database of about 2,000 moored historical observations (MHOs). The variances are computed across frequencies ranging from supertidal, dominated by the internal gravity wave continuum, to subtidal, dominated by currents and mesoscale eddies. The most important qualitative difference between HYCOM and MITgcm, and between simulations of different resolutions, is in the supertidal band, where the 1/48° MITgcm lies closest to observations. Across all frequency bands examined, the HYCOM simulations display higher spatial correlation with the MHO than do the MITgcm simulations. The supertidal, semidiurnal, and diurnal velocities in the HYCOM simulations also compare more closely with observations than do the MITgcm simulations in a number of specific continental margin/marginal sea regions. To complement the model‐MHO comparisons, this paper also compares the surface ocean geostrophic eddy KE in HYCOM, MITgcm, and a gridded satellite altimeter product. Consistent with the model‐MHO comparisons, the HYCOM simulations have a higher spatial correlation with the altimeter product than the MITgcm simulations do. On the other hand, the surface ocean geostrophic eddy KE is too large, relative to the altimeter product, in the HYCOM simulations.

Accurate forecasts of storm surges caused by winds and atmospheric pressure are important for the protection of life and property in coastal regions and also for safe navigation. Therefore, Environment and Climate Change Canada (ECCC) maintains and develops surge prediction systems. This study focuses on the assessment of the timing and amplitude of predicted surges at selected locations during the passage of hurricane Dorian. The systems use barotropic ocean models to simulate water levels and currents at 1/30 and 1/12 degree horizontal resolutions.The relatively low tidal range at the time of Dorian’s landfall helped prevent catastrophic flooding. However, with a closer superposition of the peak surge and high tide, the damage could have been more significant. Reasonably well predicted timing of the highest surge by the system helped prevent the over-issuance of warnings. Sensitivity of the forecast quality to the lead time, resolution and atmospheric forcing for the event will be presented.

Ocean General Circulation Models (OCGM) have been used for ocean forecasts and reanalysis in the past; successfully reproducing realistic large scale features such as Western Boundary Currents (WBC) which evolve slowly in time. Recent developments of three dimensional OCGM’s include the incorporation of tidal forcing embedded in the numerical integration. With the inclusion of higher frequency tidal forcing it is now possible to study the impact of tides on the larger scale features of the ocean reproduced in the OCGM’s such as WBC’s through comparison of simulations with and without tides. We compare two 1/12.5° simulations of the Hybrid Coordinate Ocean Model (HYCOM) and report on differences in the mean position, transports, and warm/cold core eddy production in the Gulf Stream and Kuroshio Current and their extensions across the Atlantic and Pacific Oceans.

Global maps of the mesoscale eddy available potential energy (EAPE) field at a depth of 500 m are created using potential density anomalies in a high‐resolution 1/12.5° global ocean model. Maps made from both a free‐running simulation and a data‐assimilative reanalysis of the HYbrid Coordinate Ocean Model (HYCOM) are compared with maps made by other researchers from density anomalies in Argo profiles. The HYCOM and Argo maps display similar features, especially in the dominance of western boundary currents. The reanalysis maps match the Argo maps more closely, demonstrating the added value of data assimilation. Global averages of the simulation, reanalysis, and Argo EAPE all agree to within about 10%. The model and Argo EAPE fields are compared to EAPE computed from temperature anomalies in a data set of “moored historical observations” (MHO) in conjunction with buoyancy frequencies computed from a global climatology. The MHO data set allows for an estimate of the EAPE in high‐frequency motions that is aliased into the Argo EAPE values. At MHO locations, 15–32% of the EAPE in the Argo estimates is due to aliased motions having periods of 10 days or less. Spatial averages of EAPE in HYCOM, Argo, and MHO data agree to within 50% at MHO locations, with both model estimates lying within error bars observations. Analysis of the EAPE field in an idealized model, in conjunction with published theory, suggests that much of the scatter seen in comparisons of different EAPE estimates is to be expected given the chaotic, unpredictable nature of mesoscale eddies.

We examine the temporal means and variability of the semidiurnal internal tide energy fluxes in 1/25° global simulations of the Hybrid Coordinate Ocean Model (HYCOM) and in a global archive of 79 historical moorings. Low‐frequency flows, a major cause of internal tide variability, have comparable kinetic energies at the mooring sites in model and observations. The computed root‐mean‐square (RMS) variability of the energy flux is large in both model and observations and correlates positively with the time‐averaged flux magnitude. Outside of strong generation regions, the normalized RMS variability (the RMS variability divided by the mean) is nearly independent of the flux magnitudes in the model, and of order 23% or more in both the model and observations. The spatially averaged flux magnitudes in observations and the simulation agree to within a factor of about 1.4 and 2.4 for vertical mode‐1 and mode‐2, respectively. The difference in energy flux computed from the full‐depth model output versus model output subsampled at mooring instrument depths is small. The global historical archive is supplemented with six high‐vertical resolution moorings from the Internal Waves Across the Pacific (IWAP) experiment. The model fluxes agree more closely with the high‐resolution IWAP fluxes than with the historical mooring fluxes. The high variability in internal tide energy fluxes implies that internal tide fluxes computed from short observational records should be regarded as realizations of a highly variable field, not as “means” that are indicative of conditions at the measurement sites over all time.

The effects of a parameterized linear internal wave drag on the semidiurnal barotropic and baroclinic energetics of a realistically forced, three-dimensional global ocean model are analyzed. Although the main purpose of the parameterization is to improve the surface tides, it also influences the internal tides. The relatively coarse resolution of the model of ~8 km only permits the generation and propagation of the first three vertical modes. Hence, this wave drag parameterization represents the energy conversion to and the subsequent breaking of the unresolved high modes. The total tidal energy input and the spatial distribution of the barotropic energy loss agree with the Ocean Topography Experiment (TOPEX)/Poseidon (TPXO) tidal inversion model. The wave drag overestimates the high-mode conversion at ocean ridges as measured against regional high-resolution models. The wave drag also damps the low-mode internal tides as they propagate away from their generation sites. Hence, it can be considered a scattering parameterization, causing more than 50% of the deep-water dissipation of the internal tides. In the near field, most of the baroclinic dissipation is attributed to viscous and numerical dissipation. The far-field decay of the simulated internal tides is in agreement with satellite altimetry and falls within the broad range of Argo-inferred dissipation rates. In the simulation, about 12% of the semidiurnal internal tide energy generated in deep water reaches the continental margins.

We apply several skill tests to assess tidal currents within a three‐dimensional, eddy resolving, global ocean circulation model compared to over 5000 observational velocity records spanning 40 years. We examine the skill of the HYbrid Coordinate Ocean Model (HYCOM) on a regional, basin, and global scale and in deep versus shallow water. On a global scale, we examine the model tidal kinetic energy (KE) compared to the tidal KE estimated from the observational velocity records. We examine the vertical structure of the model tidal KE by averaging over predetermined depth bins. We also investigate the ability of the model to satisfy the 95% confidence intervals of the individual tidal ellipse parameters. On a basin scale, we determine if any bias exists in model performance with regards to a particular part of the global ocean and further investigate if any variability of model skill exists within the ocean basins by testing the model against smaller subsets of the observations. Our results show that the skill of the nondata assimilative HYCOM is comparable to the skill of the altimetric‐constrained model TPXO7.2. HYCOM is shown to have up to 20% higher skill in resolving the Greenwich phase of the tides on a global basis and demonstrates moderate skill in replicating the vertical structure of the tidal currents as represented by the current meters. HYCOM demonstrates up to 20% higher skill than TPXO7.2 for some ocean basins and some ocean regions but exhibits up to 20% weaker skill in the Southern Ocean.