UNCERTAINTIES OF MULTI-DECADAL BUOY AND ALTIMETER OBSERVATIONS

Abstract

The impacts of climate change are evident globally. Changes to the sea state, including ocean wave heights, are important for the long-term design of oceanic infrastructure as well as resource extraction. Previous studies have estimated trends of wave height from multi-decade satellite altimetry and wave buoy observations but they do not robustly quantify uncertainties in trends. Changes to buoy hulls, payloads, and processors can introduce step changes and thus may influence year-to-year variability in the buoy time series which can create spurious climate signals. We find that standard approaches to identify step-changes in buoy time series are highly dependent on the reference time series. This is because identifying step-changes is most effective when analyzing a difference series, which is derived from the original time series minus a reference series. However, in the absence of a reliable reference series this method becomes subjective due to differences in reference series, (different hindcast products or different altimeter products). Therefore we adopt another approach which estimates uncertainties in multi-decadal wave height trends of buoy observations by extrapolating co-located altimeter-buoy wave height residuals. Synthetic wave height time series are created by randomly applying altimeter-buoy residuals to the original buoy time series through Monte Carlo simulations. The uncertainties of the buoy wave height trends are on the order of millimeters per year. This is relatively small when compared to the magnitudes of the buoy wave height trends which are on the order of centimeters per year and suggests that changes to buoy configurations are not having a large impact on the overall buoy trends. These synthetic time series provide a unique perspective into understanding wave climate and represent the uncertainty of buoy time series based on the differences between buoy and altimeters.