Effectiveness of Using Multisatellite Wind Speed Estimates to Construct Hourly Wind Speed Datasets with Diurnal Variations

Abstract

AbstractThis study used an optimum interpolation method (OIM) to construct a sea surface wind speed (SSW) dataset with hourly resolution based on merged wind speed analysis products from four satellites [AMSR-2, ASCAT, OceanSat Scatterometer (OSCAT), and WindSat]. To validate this hourly SSW dataset, the OIM dataset was compared with observations obtained from moored buoys. These buoy observations were also compared with the products of each of the four satellites individually. The root-mean-square differences and the correlation coefficients between the buoy observations and the OIM dataset indicated that the accuracy of the dataset was slightly lower than that of the single-satellite products. However, a spectrum analysis at the buoy locations indicated that the OIM dataset was capable of resolving diurnal variations in wind speed, which was a result not reproduced by the single-satellite products. In addition, the study also found that the hourly dataset with diurnal variation was effective in obtaining accurate daily mean values by reducing the sampling error. A comparison of daily mean wind speeds derived from satellite observations with those obtained from buoy observations demonstrated that greater accuracy in daily mean SSW data could be achieved using multisatellite observations in comparison with single-satellite observations. Therefore, the application of multisatellite observations that have different observation times could be a useful and effective approach with which to construct datasets with high temporal resolution and to improve the accuracy of daily mean values.