Analysis and de-noise of time series data from automatic weather station using chaos-based adaptive B-spine method

Automatic weather station (AWS) is a novel application of wireless sensor in the field of meteorological observation. The torrential data collected with AWS makes it imposable to control the quality of it in manual way. With the theory of chaos, an observed weather data series of somewhere collected by the means of automatic weather station are studied in this paper. In addition, based on the method of adaptive B-spine, we propose a novel method of denoising of observed temperature series. Comparing with traditional space means method, two experiments demonstrate better performance of the proposed way to denoise an 1-dimension chaos series.