The effectiveness of dendrogeomorphic methods for reconstruction of past spatio-temporal landslide behaviour Karel Šilhán a, ⁎, Roberta Prokešová b , Alžbeta Medveďová b , Radek Tichavský a a Department of Physical Geography and Geoecology, Faculty of Science, University of Ostrava, Chittussiho 10, 710 00 Ostrava, Czech Republic b Department of Geography and Geology, Faculty of Natural Sciences, Matej Bel University, Tajovského 40, 974 01 Banská Bystrica, Slovakia abstract article info Article history: Received 1 April 2016 Received in revised form 27 June 2016 Accepted 20 July 2016 Available online xxxx Dendrogeomorphic methods are in fact frequently used for spatio-temporal landslide behaviour reconstruction. However, their real accuracies have not yet been deeply evaluated. The purpose of this study is to evaluate the effectiveness and accuracy of tree ring based methods by comparing them with field monitoring data on land- slide behaviours. A total of 876 increment cores have been extracted from 219 P. sylvestris individuals to recon- struct the past spatio-temporal activity of the Ľubietová landslide (1977). The landslide re-activation events recorded within the tree ring series of each tree have been spatially interpolated. Those data have been compared with the long-term field monitoring based on the measured changes in the stabilized geodetic point positions. In general, the tree ring based spatio-temporal landslide evolution revealed a higher landslide activity than the monitoring-based one (particularly in the lower part of the landslide). However, higher activity was recorded from the monitoring in the northern part of the upper half of the landslide, while approximately 30% of the stud- ied landslide area revealed highly similar normalized values based on both approaches. The best fit between the results (N 80%) was in areas with mean cumulative surface displacements of approximately 200 mm (monitoring based results) and mean event recurrence time of 14.5 years (tree ring based results). The potential sources of uncertainties that should be considered in future dendrogeomorphic research are the irregular spatial positions and spatial density of trees, the physiology of the tree species, the inertia of tree growth responses within tree ring sequences and the type of studied landslide. © 2016 Elsevier B.V. All rights reserved. Keywords: Landslide Dendrogeomorphology Monitoring The Ľubietová landslide 1. Introduction Landsliding is a dangerous geomorphic process that occurs almost all over the World (Gutiérrez et al., 2010). Landslides of various types and magnitudes cause serious material damage and even fatalities every year (Van Den Eeckhaut et al., 2009) and recovery measures after landslide events can be very expensive. From this point of view, it is extremely important to realize mitigation measures on slopes with inferred landslide activity. Prediction models of landslide activity are mostly based on knowledge of past landslide spatial and temporal occurrences in a given area (Guzzetti et al., 1999; van Westen et al., 2008). Moreover, detailed data on landslide histories are important sources for landscape management planning. There are several available methods for determining landslide be- haviour. Direct field monitoring based on geodetic measurements is both very precise and very time consuming (Bányai et al., 2014; Wagner, 2016). The time span covered by that type of monitoring is mostly not longer than several decades in the case of monitored landslides. The analysis of historical orthophotos is the next possible approach, even without the need for long-term field monitoring (Prokešová et al., 2010; Ciampalini et al., 2015). The active parts of a landslide can in some cases be directly identified in orthophotos. Analyses of time series of DEMs obtained by remote sensing or direct field data collection methods are good alternatives, but such data often do not exist for studied landslides (Schlögel et al., 2015). Methods of absolute dating enable the creation of histories of land- slide activity with appropriate time resolutions, although the spatial reconstruction is limited (Lang et al., 1999). Dendrogeomorphic approaches have been presented as the most precise method of the spatio-temporal reconstruction of past landslide activity (Corominas and Moya, 1999; Stefanini, 2004; Bollati et al., 2012; Lopez Saez et al., 2012a, 2012b; Šilhán et al., 2013; Šilhán et al., 2014; Wistuba et al., 2013) with high temporal resolution (seasonal dating; Lopez Saez et al., 2012a). In fact, the methods are quite frequent- ly used for landslide activity reconstruction, but their precisions have not yet been deeply verified. Dendrogeomorphic methods are used for analyses of other geomorphic processes as well. For example, the recon- struction of past snow avalanches or flash floods require the introduc- tion of various thresholds of dendrogeomorphic indices values or the calculation of the probability of process occurrence (Butler et al., 1987; Ruiz-Villanueva et al., 2010; Corona et al., 2012). This reflects the fact Catena 147 (2016) 325–333 ⁎ Corresponding author. E-mail address: karel.silhan@osu.cz (K. Šilhán). http://dx.doi.org/10.1016/j.catena.2016.07.035 0341-8162/© 2016 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Catena journal homepage: www.elsevier.com/locate/catena