Use of Adaptive Polynomials for Real-Time Data Editing

Abstract

The National Center for Atmospheric Research has developed a new radiosonde system CLASS (Cross-chain LORAN Atmospheric Sounding System). The sounding data acquired by this system are nonstationary with time dependent mean and variance. Real-time analysis is required to determine winds, pressure, temperature and humidity. Before such computations we must carry out an editing step to remove wild data points or outliers. It is paper presents a new data-editing algorithm using adaptive polynomials. This algorithm is amenable for real-time editing of nonstationary data because of its (i) data-adaptive nature, (ii) computing efficiency, and (iii) simplicity of implementation. Its implementation is demonstrated on a set of pressure data collected by CLASS.