Assessment of the reliability of wave observations from Voluntary Observing Ships: insights from the validation of a global wind wave climatology based on Voluntary Observing Ship data

This paper describes development and validation of a global climatology of basic wave parameters based on the voluntary observing ship (VOS) data from the Comprehensive Ocean-Atmosphere Data Set collection. Climatology covers the period 1958–1997 and presents heights and periods for the wind sea, swell, and significant wave height (SWH) over the global ocean on 2° × 2° spatial resolution. Significant wave height has been derived from separate sea and swell estimates by taking square root of the sum of squares for the seas and swells propagating approximately in the same direction and assuming SWH to be equal to the higher of the two components in all other cases. Special algorithms of corrections were applied to minimize some biases, inherent in visual wave data. Particularly, we corrected overestimation of small seas, corrected underestimation of periods, and analyzed separation between sea and swell. Validation included estimation of random observational errors, observation of sampling errors, and comparison with the alternative wave data. Estimates of random observational errors show that for the majority of locations, observational uncertainties are within 20% of mean values, which allows us to discuss quantitatively the produced climatology. Biases associated with inadequate sampling were quantified using the data from high-resolution WAM hindcast for the period 1979–1993. The highest sampling biases are observed in the South Ocean, where wave height may be underestimated by 1–1.5 m because of poor sampling, primarily associated with a fair weather bias of ship routing and observation. Comparison to the other VOS-based products shows in general higher SWH in our climatology, especially in the midlatitudes. However, comparison with the altimeter data shows that even for well-sampled regions, high waves are still underestimated in VOS, suggesting a ubiquitous fair weather bias. Further ways of improving VOS-based wave climatologies and possible applications are discussed