Research On Scatterometer Data Assimilation In Wind Speed And Direction

The scatterometer is a satellite remote sensor that can provide wind speed and wind direction over the ocean near the surface. It can observe 70% of ocean and observe all day. Having the advantage of high resolution in space and temp, covering wide ocean area, which is not own by conventional ocean observation, scatterometer become the major may to observe wind over ocean. What’s more, assimilate scatterometer data effectively can make positive effect on ocean weather forecast.The research starts with the retrieval of scatterometer data. Next, do some study about pre-progress of observation data. Then, as there are some problems in data assimilation of scatterometer data, I give a new thought that assimilate scatteromerter in speed and direction, and design two speed and direction operator for 10 m real wind and 10 m neutral wind.Backscatter is related to the roughness of ocean, which is decided by wind over ocean near the surface. So winds over the ocean near surface can retrieval by scatterometer data. Geophysical Model Function(GMF) describes the function between wind(speed and direction) and backscatter. Given scatterometer data, we can get the wind relatively through GMF. MLE method that is based on Bayes Theory is the common retrieval method. According to different GMF, scatterometer data can be retrieval as 10 real wind and 10 neutral wind.The retrieval wind of scatterometer gives several ambiguity wind direction. When we use the retrieval wind, we must choose a solution of ambiguity wind direction. We call this progress direction ambiguity removal. 2DVAR is a advanced ambiguity removal method, which use Fourier transformation, Helmholz transformation and Normalization in cost function. Through transformation of cost function, it is easy to calculate the B and B-1. Scatterometer data have bias. Triple location method, that use buoy wind and NWP wind and NWP wind to correct scatterometer wind. The corrected scatterometer wind has a zero value of the first moments of the random measurement errors.Although most wind observation system observe wind speed and direction, the wind data is transformed to u, v before assimilation. Thus, assimilate result is independent of observe direction error. If assimilating wind in speed and direction directly, observe direction error can effect assimilate result. So I design speed and direction operator of real wind, and speed and direction operator of neutral wind.Finally, I do assimilating experiments with speed and direction operator of real wind. The Result shows this new operator is better than traditional operator. Besides, the Value of speed and direction operator of neutral wind is needed to do some experiments.