FCXNL 09A02 1652072 1 asbt2 gulev by C90fCqm


									Visual wind wave data from VOS: a substantial component of wind wave observing

Sergey K. Gulev and Vika Grigorieva (P.P. Shirshov Institute of Oceanology, RAS,

Visual wave observations (assimilated in ICOADS) are available effectively from the mid 19th century
and represent the longest records of wind wave information worldwide taken with a unique
observational practice. Visual wave data are characterized by quite strong systematic and random
errors. Maximum random observational errors in wins sea height amount to 1 meter with maximum
random observational errors in swell height being up to 1.6 meters. Significant uncertainties (both
random and systematic) in wind wave periods estimates of up to 2-3 seconds may result from the
deviation of the actual observational practices from the guidelines. VOS-based climatological
estimates of wave characteristics also suffer from spatially and temporally inhomogeneous sampling
with the largest sampling errors (up to 1.5-2 meters in wave height) identified in the poorly observed
regions of the Southern Ocean and subpolar Northern Hemisphere. We will present 60-yr climatology
of wind waves based on visual observations. It includes estimates of heights and periods of wind sea
and swell as well as derived SWH and dominant period along with error estimates. Climatology
allows for the analysis of linear trends and patterns of interannual variability in wind wave
characteristics worldwide. Analysis of the earlier 20th century records of visual wave data for selected
ship routes demonstrated centennial increase of SWH (of 8-10 cm/decade) in the North Pacific, with
no significant centennial trends in the Atlantic. Visual VOS data also allow for estimates of extreme
wind waves for the last several decades, if only in well sampled North Atlantic and North Pacific mid
latitudes. Further prospects for the improvement of the accuracy of visual wind wave data will be
discussed. These include installation onboard of selected VOS rolling sensors and recording of ship
radar scans for providing alternative data for the validation of visual wave estimates.

To top