A Statistical method for estimating the convective mixing height

Abstract

An important parameter often required by air quality models is the so called mixing height. In day time hours (convective situations), three over- lapped layers may be distinguished in the lowest part of the atmosphere: the superficial layer, the mixed layer and the entrainment layer. The latter is a thin atmospheric layer that reverses temperature and buffers the up- per free atmosphere. Commonly, atmospheric variables are discontinuous at the entrainment layer. In this approach this latter fact is used to develop a change point analysis for atmospheric measurements. Then, the change point considered in the statistical model is precisely the mixing height. We also discuss the a priori distributions for the parameters in our model and the techniques needed to perform the necessary calculations. Some examples are analysed using data obtained in Texcoco, Mexico.