A new planetary boundary layer height estimate method based on wavelet and random forest

In order to overcome the shortcomings of common lidar boundary layer height estimation methods in complex atmospheric conditions such as clouds and suspended aerosols, a planetary boundary layer height estimation method based on wavelet and random forest is proposed. First, the boundary layer height estimation model based on the random forest is built from lidar data set. Then, the wavelet covariance method and the model are used to estimate the boundary layer height respectively. Last, the final value is obtained from wavelet covariance by reference the output value of the model. The experimental results show that, compared with the random forest method and wavelet covariance method, the boundary layer height estimated by the proposed method is close to the radiosonde measurement with very small error. It can be seen that the proposed method is a high performance atmospheric boundary layer height estimation method for complex atmospheric conditions.