Abstract The spatial and temporal variability of ocean wind waves plays an important role in many engineering and environmental problems. Although research in this area has been improved in recent decades thanks to the emergence of satellite data, in many cases this information does not have the appropriate resolution for more detailed and local research. In view of this, reanalysis data developed by several meteorological agencies have appeared as a good alternative to force the most popular ocean wind-wave models. Thus, to achieve more accuracy in the data, the 60-yr Global Atmospheric Reanalysis 1 carried out by the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) has been corrected, employing the vector correlation and triple collocation theories combined with information from different accurate sources. The comparison of wind speed with satellite and in situ buoy data before correction reveals an important underestimation for areas near the Colombian coast. The wind speed root-mean-square error (RMSE) between corrected data and satellite measurements at locations near the Colombian Caribbean coast without calibration is 3.8 m s−1, while for corrected data it is 2.0 m s−1, showing a decrease in the RMSE of almost 47%. For significant wave height for buoy 41194 (Barranquilla, Colombia), the RMSE between the modeled data and measurements without correction is 0.99 m, while for the corrected data it is 0.40 m, showing a decrease in the RMSE of almost 55%. The results obtained clearly show an increase in the accuracy of the calibrated wind speed.