(Multiple) linear regression (MLR) and selected nonlinear methods from the field of machine learning were compared for the analysis of relationships between xylem tree-rings and the environment: artificial neural networks with a training algorithm that uses Bayesian regularization (ANN), model trees (MT), ensembles of model trees (BMT) and random forests of regression trees (RF). The selected methods were compared on nine datasets, which included different tree-ring parameters and different target environmental variables. For the nonlinear methods, better statistical metrics were calculated on validation data in most cases, but the differences in comparison to linear regression were minor. Additional analysis indicated that the methods mostly differ in predicting the extreme values. The characteristic of nonlinear methods is that the change in the dependent variable is not proportional to the change of one or more independent variables. The latter results in a reduced range and variability of reconstructed values, which makes the reconstruction visually less attractive as compared to linear extrapolation, even though in most cases statistically better. None of the nonlinear machine learning methods showed best results on all datasets, therefore it makes sense to always compare different machine learning regression methods prior to climate reconstruction. To do so, the R function compare_methods() was developed and implemented in the dendroTools R package, which is freely...
D.09 Tutoring for postgraduate students
COBISS.SI-ID: 930167This master thesis presents the use of dendrochronology as a method of archaeological natural sciences for understanding the climate in the past. Samples of living trees and historical/archaeological wood in Krško Basin and Lonja Field were taken and afterwards analysed in a dendrochronological laboratory. The author presumed the samples would allow for an extension of the reference chronologies of Krško Basin and Lonja Field. This would serve as the basis for the long reconstruction of the climate and the unification of the chronologies into a single climatic model of western Pannonian Plain. The Krško Basin climate was reconstructed till the year 1669, while that of Lonja Field reached the year 1560. Using PDSI JJA, the analysis of pointing years was made, which found no correlation between the two areas. The author explains this by showing the differences in the geography of Krško Basin and Lonja Field. Written historical sources that correspond with identified pointing years were also identified though those regarding Lonja Field exceeded those for Krško Basin. For a better understanding of the chronologies and their extension even further into the past continued sampling is required.
D.10 Educational activities
COBISS.SI-ID: 70749794In this thesis, the following three objects from the Lower Carniola have been dated by dendrochronology: the Škarjat's hayrack from Mirna, the church of Saint Vid from Šentvid near Stična, and the Kosmo's wine cottage from Mali Cirnik. All of them are made of oak wood (Quercus sp.) and compared to Lower Carniola's chronology of the Slovenian Forestry Institute. The dating was done with the help of the statistical analyses - the correlation analyses, t values by Baillie-Pilcher and Gleichläufigkeit, and the visual comparison of curves with reference to chronologies. It was impossible to date the church of Saint Vid with the standard comparison methods. Because it is so old, the C14 analyses were made instead. Hayrack from Mirna was dated after the year 1626, wine cottage after 1616 and the church somewhere after 1483. The analysed church's samples were parts of its renovation not the original construction, because the church's Romanesque base was renovated in the Gothic style after 1483.
D.10 Educational activities
COBISS.SI-ID: 38176771We introduce in this paper the dendroTools R package for studying the statistical relationships between tree-ring parameters and daily environmental data. The core function of the package is daily_response(), which works by sliding a moving window through daily environmental data and calculating statistical metrics with one or more tree ring proxies. Possible metrics are correlation coefficient, coefficient of determination and adjusted coefficient of determination. In addition to linear regression, it is possible to use a nonlinear artificial neural network with the Bayesian regularization training algorithm (brnn). dendroTools provides the opportunity to use daily climate data and robust nonlinear functions for the analysis of climate-growth relationships. Models should thus be better adapted to the real (continuous) growth of trees and should gain in predictive capabilities. The dendroTools R package is freely available in the CRAN repository. The functionality of the package is demonstrated on two examples, one using a mean vessel area (MVA) chronology and one a traditional tree-ring width (TRW).
F.23 Development of new system-wide, normative and programme solutions, and methods
COBISS.SI-ID: 5178534