Exploiting the structure of MWR-derived temperature profile for stable boundary-layer height estimation

摘要

A method for the estimation of Stable Boundary Layer Height (SBLH) using curvature of the potential temperature profiles retrieved by a Microwave Radiometer (MWR) is presented. The vertical resolution of the MWR-derived temperature profile decreases with the height. A spline interpolation is carried-out to obtain a uniformly discretized temperature profile. The curvature parameter is calculated from the first and second order derivatives of the interpolated potential temperature profile. The first minima of the curvature parameter signifies the point where the temperature profile starts changing from the stable to the residual conditions. The performance of the method is analyzed by comparing it against physically idealized models of the stable boundary-layer temperature profile available in the literature. There are five models which include stable-mixed, mixed-linear, linear, polynomial and exponential. For a given temperature profile these five models are fitted using the non-linear least-squares approach. The best fitting model is chosen as the one which fits with the minimum root-mean-square error. Comparison of the SBLH estimates from curvature-based method with the physically idealized models shows that the method works qualitatively and quantitatively well with lower variation. Potential application of this approach is the situation where given temperature profiles are significantly deviant from the idealized models. The method is applied to data from a Humidity-and-Temperature Profiler (HATPRO) MWR collected during the HD(CP)~2 Observational Prototype Experiment (HOPE) campaign at Jiilich, Germany. Radiosonde data, whenever available, is used as the ground-truth.