掩星接收机误差对大气温度反演精度影响仿真研究

为了评估掩星接收机误差对无线电掩星探测精度的影响,利用EGOPS软件仿真研究了多普勒偏差、多普勒频移、时钟稳定性/单差分、接收机噪声和局部多路径等对GNSS无线电掩星反演大气温度、折射率、密度、压力廓线的影响.模拟设计过程中选择METOP作为接收的LEO卫星,GPS星座作为发射系统,设置GRAS天线.从模拟的567个掩星事件中选择了一个上升掩星事件和一个下降掩星事件进行模拟分析.研究结果表明:对于设置的"最差"情况,温度误差最大值大部分出现在平流层顶附近,其中多普勒偏差引起的温度误差最大值接近2 K,多普勒频移引起的温度误差最大值小于0.3 K,时钟稳定性/单差分引起的温度误差最大值接近3 K,现实接收机噪声引起的温度误差最大值超过了4 K,局部多路径引起的温度误差最大值小于1.5 K.经讨论分析认为:对于高质量的无线电掩星反演,掩星接收机误差源中最主要的是接收机热噪声、时钟稳定性/单差分和多普勒偏差.文中还随机选取了30个掩星事件进行统计分析,将其温度反演结果与ECMWF分析场数据比较得出,温度误差的平均偏差在45 km高度上最大,约为1 K;最大标准偏差出现在平流层顶,约为5 K,这验证了廓线反演算法的有效性,表明了误差分析结果的正确性和普遍性.

Simulation study on temperature retrieval deviations induced by occultation receiver errors

In order to evaluate the influences of occultation receiver errors on the radio occultation sounding accuracy, we employ the EGOPS simulation software to evaluate the influences of Doppler biases and drifts, clock stability/single-differencing, receiver thermal noises and local multi-path on the retrievals of temperature, refractivity, density and pressure profiles. The METOP and GPS constellation are selected as the LEO satellite and transmit system respectively, and GRAS antenna is setting in the simulation process. From the 567 simulated occultation events, one rising and one setting occultation events are tested in the simulation. The results show that the maximum temperature deviation is about 2K due to the influence of Doppler bias, 0.3K due to the influence of Doppler shift, less than 1.5K due to the influence of local multi-path, 3K due to the influence of Clock stability/single-differencing, and over 4K due to the influence of receiver thermal noises, respectively. It can be seen that the worst situations for temperature retrieval occur near the stratopause. The major error sources are receiver thermal noises, Clock stability/single differencing and POD-induced Doppler biases. For the purpose of statistical analysis, thirty occultation events are randomly selected, and their temperature retrievals are compared with corresponding ECMWF data. We find that the average deviation of retrieved temperature errors reaches its maximum, 1K, at 45Km height, and the standard deviation its maximum, 5K, at the stratopause.