资源环境科技发展态势分析平台

Satellites provide the most extensive surface wind data over the tropical oceans, but sampling limitations, rain contamination, and algorithm uncertainties remain. In this study, scatterometer and passive microwave swath data, and multisatellite products are evaluated using Dynamics of the Madden-Julian Oscillation (DYNAMO) ship and moored buoy wind speed data over the Indian Ocean from September 2011 to January 2012. Each of the satellite products has <2 m/s error, consistent with previous studies. Blended products have error and bias characteristics comparable with swath data and provide better coverage. The errors and biases depend on the prevailing weather regime. The equatorial region is characterized by weeks of winds <5 m/s with little rain punctuated by episodes of winds >10 m/s and heavy rainfall associated with MJO events. The blended products resolve the episodic enhanced equatorial winds associated with the active phase of the MJO. However, due to limited sampling and rain contamination, the swath data did not fully resolve the peak MJO winds. In the Southern Hemisphere trade wind regime, rain contamination is less of a factor and winds are steadier. Satellite wind estimates of equatorial (trade wind) regime tended to be biased low (high). The surface wind products underestimate higher winds during the active MJO due to limited sampling and have high wind bias in the trade wind regime. These weather regime-dependent biases should be taken into account for applications using these wind products for ocean circulation studies and estimating air-sea fluxes over the Indian Ocean.