基于核函数NPSSB模型的HY-2A卫星雷达高度计海况偏差研究

利用机器学习的方法,对14个周期HY-2A卫星高度计数据:风速、有效波高和海面高度差值进行训练,探究海况偏差和风速、有效波高之间的关系,创建海况偏差核函数非参数模型(NPSSB),并与参数模型中具有代表性的BM3、BM4模型进行对比。研究表明:(1)核函数NPSSB模型能够很好的反映SSB与U、SWH之间的关系,SSB与U呈二次函数关系,SSB与SWH呈反比例函数关系;(2)核函数NPSSB模型对SSB的模拟能力与训练数据集相关,数据量越多,模拟能力越好;(3)核函数NPSSB模型与BM3、BM4模型都存在0～-0.03 m的差值,随着风速和有效波高的增加,差值的绝对值越大。

Study on the sea state bias of HY-2A satellite radar altimeter based on a nonparametric kernel function model

Based on machine learning method, this paper investigates the relationship between sea state bias and wind speed and significant wave height by training wind speed, significant wave height and sea surface height difference observed by 14 cycles of HY-2A satellite altimeter. Meanwhile, a nonparametric kernel function model for sea state bias(NPSSB)is established and is compared with typical parametric model BM3 and BM4. The results show that the relationship between SSB and wind speed and SWH is well represented by NPSSB model.SSB shows quadratic function and inverse proportion function relationship with U and SWH, respectively.(2) The ability of NPSSB model in simulating SSB is related to the amount of training dataset. The more the data is, the better the simulation ability will be. The difference between NPSSB model and the BM3 and BM4 models is both between-0.03 m and 0, the absolute value of which increases with the increase of wind speed and SWH.