Journal cover Journal topic
Climate of the Past An interactive open-access journal of the European Geosciences Union
Clim. Past, 12, 1215-1223, 2016
https://doi.org/10.5194/cp-12-1215-2016
© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.
Technical note
24 May 2016
Technical note: Estimating unbiased transfer-function performances in spatially structured environments
Mathias Trachsel1,a and Richard J. Telford1,2 1Department of Biology, University of Bergen, PO Box 7803, 5020 Bergen, Norway
2Bjerknes Centre for Climate Research, Bergen, Norway
acurrently at: Department of Geology, University of Maryland, College Park, MD 20742, USA
Abstract. Conventional cross validation schemes for assessing transfer-function performance assume that observations are independent. In spatially structured environments this assumption is violated, resulting in over-optimistic estimates of transfer-function performance. H-block cross validation, where all samples within h kilometres of the test samples are omitted, is a method for obtaining unbiased transfer-function performance estimates. In this study, we assess three methods for determining the optimal h. Using simulated data, we find that all three methods result in comparable values of h. Applying the three methods to published transfer functions, we find they yield similar values for h. Some transfer functions perform notably worse when h-block cross validation is used.

Citation: Trachsel, M. and Telford, R. J.: Technical note: Estimating unbiased transfer-function performances in spatially structured environments, Clim. Past, 12, 1215-1223, https://doi.org/10.5194/cp-12-1215-2016, 2016.
Publications Copernicus
Download
Short summary
In spatially structured environments, conventional cross validation results in over-optimistic transfer function performance estimates. H-block cross validation, where all samples within h kilometres of the test samples are omitted is a method for obtaining unbiased transfer function performance estimates. We assess three methods for determining the optimal h using simulated data and published transfer functions. Some transfer functions perform notably worse when h-block cross validation is used.
In spatially structured environments, conventional cross validation results in over-optimistic...
Share