Profiling boundary layer temperature using microwave radiometer in east coast of china

Abstract

The boundary layer temperature profile is very essential for modeling atmospheric processes, whose information can be obtained using radiosonde data generally. Beside this, ground-based multichannel microwave radiometer (GMR) offers a new opportunity to automate atmospheric observations by providing temperature, humidity and liquid water content with high time resolution, such as MP-3000A ground-based multi-channel radiometer. An experiment in east coast of China for profiling boundary layer temperature was performed at Qingdao Meteorological Station from 1 March to 23 April in 2014 using an MP-3000A radiometer. Three techniques have been applied to retrieve the boundary layer temperature profile by using the experimental data, namely the linear regression method, the back propagation (BP) neural network method and the 1-D Variational (1D-VAR) method. Elevation scanning is introduced to help improve the accuracy and resolution of the retrievals for each technique. These results are compared with the radiosonde data at the same time. The preliminary results achieved by each method show that the average day root-mean-square (rms) error for temperature is within 1.0 K up to 2 km in height. The 1D-VAR technique seems to be the most effective one to improve the precision of the boundary layer temperature profile.