Finding a proxy for wind stress over the coastal ocean

J. E. Wood*, M. Roughan, P. M. Tate

*Corresponding author for this work

    Research output: Contribution to journalArticlepeer-review

    24 Citations (Scopus)


    Wind stress forcing is a critical driver of oceanographic processes. In the absence of over-ocean wind measurements, re-analysis products and over-land measurements are often used. The present paper compares a unique wind-data series from an ocean mooring with two re-analysis products and data from six over-land sites in Sydney, Australia, to determine whether these data can infer over-ocean wind conditions. Four oceanographic moorings are located here; however, over-ocean meteorological observations are no longer available. Correlations between over-ocean and over-land sites wind stress were > 0.8, whereas for re-analysis products, correlations ranged from 0.28 to 0.72. Re-analysis products were unable to resolve variability at the over-ocean site with periods shorter than 2 days, indicating that they are not appropriate wind proxies for the coastal ocean. Somewhat counter-intuitively, our results showed that proximity of over-land sites to the over-ocean site does not necessarily imply the highest correlation or the lowest error and careful site selection is required. An upwelling study showed that wind observations from selected over-land sites can accurately represent isotherm uplift in the coastal ocean. Threshold wind stress values that uplift isotherms at these over-land sites are provided, along with a recommendation for the use of the over-land site that best represents over-ocean wind conditions in this region.

    Original languageEnglish
    Pages (from-to)528-544
    Number of pages17
    JournalMarine and Freshwater Research
    Issue number6
    Publication statusPublished - 2012


    • IMOS
    • Integrated Marine Observing System
    • New South Wales
    • NSW


    Dive into the research topics of 'Finding a proxy for wind stress over the coastal ocean'. Together they form a unique fingerprint.

    Cite this