Nonlinear Regression Techniques for Estimation of Missing Values in Daily Temperature Series

Abstract:

While time series of meteorological measurements from land-based weather stations still represent one of the basic types of data employed in the climate research, it not uncommon for these records to be incomplete, interrupted by shorter or longer periods of missing values. Such gaps often need to be filled before a subsequent analysis can be performed, and records from other nearby measuring sites are frequently used for this purpose. In this presentation, results of central European daily mean, minimum and maximum temperature estimation from other concurrent temperature measurements by various statistical methods are presented, with a particular emphasis on assessing potential benefits of nonlinear regression techniques application.

Using multi-decadal daily temperature series originating from a dense network of weather stations covering the territory of the Czech Republic, we show that while nonlinear regression does not always outperform its linear counterpart, it can substantially improve accuracy of temperature estimates for some target locations. The gain is especially prominent for sites exhibiting atypical behavior compared to their local geographic neighborhood (such as isolated mountain-based stations). Performance of various regression techniques is also compared to selected geostatistical interpolation methods, and suitability of different approaches to missing data estimation is discussed.