高斯函数在香港地区对流层层析实验中的应用

The Application of Gauss Function in Tropospheric Tomography in Hong Kong Area

将GPS信号的斜路径湿延迟当作层析的观测量能够有效获取对流层的三维水汽场。由于射线分布的不均匀和观测网地形的扁平,观测方程是不适定的,因此需要添加一些约束条件来确定唯一解。由于水汽在垂直方向变化很快,合理的垂直约束在获取准确的水汽场上起着重要作用。研究了香港地区湿折射率的垂直分布特征,发现高斯函数能很好地表达湿折射率与高度的关系,利用高斯函数建立约束方程获得的层析解能很好地与探空数据和欧洲中尺度天气预报中心(ECMWF)数据吻合。相对于指数约束所得结果,层析湿折射率的标准差在整个对流层减小了3.8mm/km,在低对流层减小了4.7mm/km。实验也表明,利用其他气象数据,如无线电探空数据,作为湿折射率的先验信息,也可以得到较好的层析解。

In recent three decades,the development of Global Positioning System（GPS）highly extends its application fields,one of which is to use GPS monitor water vapor in the troposphere.When Slant wet delays of GPS signals are treated as tomography observations,they can be used to retrieve three-dimensional wet refractivity fields of the troposphere.However,due to the uneven distribution of the signal rays and flat orography of the network in Hong Kong,the tomography observation equations are ill-posed,so some constraints are usually added to determine the unique tomography solution.since the water vapor changes rapidly in the vertical direction,appropriate vertical constraints play an important role in retrieving the accurate vertical structure of the wet refractivity fields.By investigating the vertical distribution characteristics of the wet refractivity in the atmosphere in Hong Kong area,we find that the Gaussian function could well express the relationship between the wet refractivity and the height.The tomographic experiments using data from Hong Kong Satellite Positioning Reference Station Network shows that tomographic solution using Gaussian function to establish vertical constraints gets a better agreement with the radiosonde data and the European Centre for Medium-Range Weather Forecasts（ECMWF）data when compared with the solution obtained by establishing vertical constraints by using the exponential function.The improvements with respect to standard deviation are 3.8mm/km in the whole troposphere and 4.7mm/km in the lower troposphere.The experiments also show that using good a priori wet refractivity from other meteorological data sources,like radiosonde,could help obtain good tomographic solutions.