Impact of using multiple-satellite sensors on the accuracy of daily-mean sea surface wind data

Abstract

ABSTRACT Most satellites observing sea surface winds have sun-synchronous orbits and provide observation data at the same place two times per a day. A daily mean value estimated from these data suffers from sampling errors because the high-frequency variation including a diurnal change in the wind field cannot be neglected. To overcome this problem, the use of multiple satellites is useful. The purposes of this study were to describe the time variation of the accuracy of daily mean wind data in a third-generation Japanese Ocean Flux Data Sets with Use of Remote Sensing Observations (J-OFURO3) and to investigate its causes by comparison with in situ measurement data from moored buoys. The results reveal that the three statistical measures such as bias, Root Mean Square error and cross correlation coefficient have improved over time. A set of scatter diagrams of the number of satellites versus the statistical measures for each year shows strong correlations. The accuracy of daily mean data provided by J-OFURO3 is concluded to depend on the number of satellites. We also focused on specific time intervals for satellite wind observations, particularly the maximum missing time interval (MMTI) within a day. The results showed that the correlations between the three statistical measures and MMTI were quite high. Because the above-mentioned two causes are not independent, we analysed the combined effect of the two causes together. The results show that the accuracy of daily mean data depends more strongly on MMTI than on the number of observations. By applying these results, it is possible to determine the optimal number of sensors and the optimal observation time to achieve maximum target accuracy. This provides very useful information for the design of satellite observation systems for sea surface wind.