抄録
To clarify the characteristics of various global sea surface wind speed (SSW) products, we evaluated SSW data in the products by using those observed by 103 moored buoys. From this evaluation, we found that the SSW data of Quick Scatterometer (QuikSCAT) without using rain flag and Japanese Ocean Flux Data sets with Use of Remote Sensing Observations version 2 (J-OFURO2) have relatively large positive bias compared with other products. Because the Cross-Calibrated Multi-Platform (CCMP) has the smallest root mean square (RMS) error and the largest correlation coefficient in this evaluation, we concluded that the CCMP gives the best performance. Thus, we carried out inter-comparisons between the CCMP and each of the remaining 14 products. SSW data of QuikSCAT and J-OFURO2 have large positive mean differences for the CCMP SSW data, and local maxima of RMS differences can be found over the inter-tropical convergence zone (ITCZ) and the South Pacific convergence zone (SPCZ). Moreover, we investigated the cause of overestimation by QuikSCAT and found the overestimation to be caused by rain contamination included in the QuikSCAT SSW data. However, if we carry out quality control for the QuikSCAT SSW data by using rain flag information, not only positive bias but also RMS error for QuikSCAT is effectively reduced. Therefore, the QuikSCAT SSW data without rain-contaminated data have relatively good accuracy. In addition, the CCMP properly represents an orographic effect in the coastal regions because of inclusion in the model-derived product, and the CCMP has the best accuracy in the coastal regions as well as in the open ocean.
