Improvement and Implementation of the Homogeneous Segment Merging Technique for Climate Datasets

University Of Alabama In Huntsville, Huntsville AL

Investigators

Abstract

This project will address the issue of inhomogeneity in the observed climate record in several regions of the U.S., within 45 km of Huntsville, Alabama and Central California. The method developed by the principal investigator makes use of homogeneous segments of the surface air temperature data record, taking into account the re-location of stations, and merging them into a single, homogeneous time series for the Huntsville area. This method is also to be applied to data sets at different altitudes in California. The approach will provide mathematical rigor to a task which is often in danger of being subjective, employing a matrix reduction method already developed by the P.I. A least-squares fit to all the entries using an interative/cumulative technique will generate the most probable, non-unique solutions. The results of this work will be submitted to major scientific journals and conferences. The mathematical method developed for this work (aimed at reducing biases between proximate data stations) is expected to have applications to variables other than temperature. The resulting records will add clarity to the extent of changes in several localized temperature records spanning the 20th century.