It is rare that there exists a sufficiently long term history of accurate measurements of extreme storm generated winds, waves, currents and storm surge to estimate the probability distribution of extremes in the range of rare probabilities needed for calculation of design loads on offshore or coastal structures. At a few sites, instrumented platforms or moored buoys have acquired data over the past twenty years or so, and satellite altimeters have measured global wave heights over about a ten-year period. However, while such data make it technically possible to compute extremes directly from the measurements, the reliability of such estimates must be questioned at least on the grounds that natural climate variability on decadal time scales is not properly represented. Therefore, even in areas with measured data it is still advisable to generate the long term data base needed to estimate the climate of extremes through a hindcast approach. Oceanweather's principals pioneered the development and application of this approach during their tenure at New York University in the 1960s and 1970s and extensively refined and applied this approach to the many new offshore development projects launched within the past two decades (see the recent review by Cardone,1999).

Briefly, the hindcast method as applied to the specification of the extreme climate consists of the following steps: (1) survey of historical meteorological data over a period of typically 50 years or longer, to identify the most severe storms of the relevant type or types responsible for extremes; (2) for each storm selected above a threshold of intensity, numerical hindcast of the time history of the sea state and currents on a grid of points representing the basin; (3) calculation of the expected extreme wave heights and associated properties, and currents for each storm at each point; (4) extrapolation of the hindcast and calculated extremes through the process of extremal analysis, which provides estimates of extremes associated with specified return periods (return period is the average interval in years between events equal to or greater than the associated extremes). Such data are then used by ocean engineers for the specification of design loads on structures.

The hindcast of an individual historical storm consists of two basic steps. First, the time and space evolution of the surface marine wind field must be specified as accurately as possible, a process which usually requires the reanalysis of historical meteorological data by experienced Meteorologists with the aid of calibrated objective analysis procedures and/or models. The wind fields are used to drive calibrated ocean response models (e.g. a spectral wave model, storm surge model, ocean current model) as the second part of the process.

The prototype for modern hindcast studies was the Gulf of Mexico Ocean Data Gathering Program (ODGP), which began in 1969 and included an extensive measurement program, a wave hindcast model development and calibration phase and a hindcast phase, all culminating in the establishment of reliable extreme wave heights and wave periods associated with hurricane generated sea states in deeper parts of the Gulf of Mexico continental shelf between the Mississippi Delta and the Texas/Mexico border (Cardone et al., 1976, Ward et al., 1979; Haring and Heideman, 1978).

Since ODGP, hundreds of dedicated hindcast studies have been carried out at Oceanweather for the offshore and coastal engineering communities to develop reliable extreme wind, wave, surge and current design estimates for design of specific offshore and coastal structures. Within the past decade many basin-wide hindcast studies have been supported jointly by many operators. These so-called Joint Industry Projects, or JIPs, have also included the application of the hindcast approach as described above to the hindcast of several continuous years for the specification of the operational climate. Major JIPs administered and/or carried out by Oceanweather within the past decade have addressed the Gulf of Mexico (GUMSHOE ad WINX), the Bering Sea (BSCOMP) Chuckchi Sea (CSCOMP), Brazil (BOMOS), Russian Arctic Seas (RASMOS), Sakhalin Island (SIMOS), South China Sea (SEAMOS), Nile Delta of Egypt (Nile Delta), Caspian Sea (CASMOS), North Sea (NESS/NUG/NEXT), West Africa (WAX/WANE) and the Caribbean Sea (CARIMOS). Major new continuous hindcasts by Oceanweather include a 40-year hindcast of the North Atlantic (AES40, Swail and Cox, 2000) and the entire globe (GROW, Cox and Swail, 2001).

The skill of storm hindcasts that is necessary in order for a hindcast database to provide extrapolation of reliable sea state extremes for design has been demonstrated and achieved in validation studies carried out as part of the studies and JIPs noted above. A recent review is given in Cardone et al. (2000). As first demonstrated for a wide range of tropical cyclones and severe extratropical cyclones by Reece and Cardone, (1982) the hindcast method may typically specify peak significant wave height (SWH) at an arbitrary site in an historical storm with bias of less than 0.5 m, mean absolute error of less than 1.0 m and scatter index (SI) of 10-15% (SI= 100 x sd/avg where sd is the standard deviation of differences between hindcast and measured peak wave heights and avg is the average of measured heights in the validation population of heights; SI is also often expressed fractionally as sd/avg). The peak spectral period (TP) appeared to be specified with comparably small bias but with greater scatter. Cardone et al., (1995) also demonstrated that where surface wind fields are specified using kinematic reanalysis techniques which take advantage of the enhanced data coverage in areas of dense buoy and/or offshore platform measurement arrays (e.g. off the east and west coasts of North America and in and around the North Sea), well calibrated wave models may specify the evolution of SWH with negligible bias and scatter near the lower limit set by accuracy and sampling variability in the wave measurements not only in storms but for continuous periods as well. As the density and quality of wind observations decreases so does the accuracy of the analyzed wind fields and the skill of ocean response hindcasts. However, the validation of Oceanweather's new long term wave hindcast of the global oceans (GROW) indicate that errors in hindcasts validated against wave measurements from buoys and satellite altimeters are acceptably low for many design purposes (but not 100-year extremes) with SWH SI of less than 25% and bias of less than 0.5 m (Cox and Swail, 2001). Recent validation studies have added hindcast-measured parameter distributional comparisons as a skill measure, usually in terms of SWH quantile-quantile scatter plots (Cox et al., 1999).

REFERENCES

1976. Cardone, V. J., W. J. Pierson and E. G. Ward. Hindcasting the directional spectra of hurricane generated waves. J. Petrol. Technol. 28, 385-394.

1979. Ward, E. G., L. E. Borgman and V. J. Cardone. Statistics of hurricane Waves in the Gulf of Mexico. OTC 3229. 10th Annual Offshore Technology Conference, 8-11 May, 1978, Houston, TX. (also appeared in J.Petrol. Tech., 1979)

1980. Haring, R.E. and J. C. Heideman. Gulf of Mexico rare wave return periods. J. Petrol. Tech ., 35-37.

1982. Reece, A. M. and V. J. Cardone. 1982. Test of wave hindcast model results against measurements during four meteorological systems. Offshore Technology Conference. OTC 4323, 269- .
1995. Cardone, V. J., H.C. Graber, R.E. Jensen, S. Hasselmann and M. J. Caruso. In search of the true surface wind field in SWADE IOP-1: ocean wave modeling perspective. The Global Atmosphere and Ocean System. 3. 107-150.

1999. Cardone, V. J. Predicting the extremes of marine winds and waves. Summary of presentation at MetOcean Workshop on Deep Water and Open Oceans, February 21, 1999, Houston, TX. Sponsor: Society for Underwater Technology.

1999. Cox, A.T., V. J. Cardone and V. R. Swail. On the use of in-situ and satellite wave measurements for evaluation of wave hindcasts. CLIMAR 1999 Preprints. Sept. 8-15, 1999, Vancouver, B.C.

2000 Swail, V.R. and A.T. Cox. On the use of NCEP/NCAR Reanalysis Surface Marine Wind Fields for a Long Term North Atlantic Wave Hindcast. J. Atmo. Tech., Vol. 17, No. 4, pp. 532-545

2000 Cox, A.T. and V.R. Swail. A Global Wave Hindcast over the Period 1958-1997: Validation and Climate Assessment. J. of Geophys. Res. (Oceans) Vol. 106, No. C2, pp. 2313-2329. (Oceans).