KEY MESSAGE: The machine learning algorithm extreme gradient boosting can be employed to address the issue of long data gaps in individual trees, without the need for additional tree-growth data or climatic variables. ABSTRACT: The susceptibility of dendrometer devices to technical failures often makes time-series analyses challenging. Resulting data gaps decrease sample size and complicate time-series comparison and integration. Existing methods either focus on bridging smaller gaps, are dependent on data from other trees or rely on climate parameters. In this study, we test eight machine learning (ML) algorithms to fill gaps in dendrometer data of individual trees in urban and non-urban environments. Among these algorithms, extreme gradient boosting (XGB) demonstrates the best skill to bridge artificially created gaps throughout the growing seasons of individual trees. The individual tree models are suited to fill gaps up to 30 consecutive days and perform particularly well at the start and end of the growing season. The method is independent of climate input variables or dendrometer data from neighbouring trees. The varying limitations among existing approaches call for cross-comparison of multiple methods and visual control. Our findings indicate that ML is a valid approach to fill gaps in individual trees, which can be of particular importance in situations of limited inter-tree co-variance, such as in urban environments. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s00468-024-02573-y.

Department of Geography Johannes Gutenberg University Johann Joachim Becher Weg 32 55128 Mainz Germany

Global Change Research Centre Brno Czech Republic

Zobrazit více v PubMed

Aryal S, Häusser M, Grießinger J, Fan Z, Bräuning A (2020) “dendRoAnalyst”: A tool for processing and analysing dendrometer data. Dendrochronologia 64:125772. 10.1016/j.dendro.2020.125772

Barrao S, Serrano-Notivoli R, Cuadrat JM, Tejedor E, Saz Sánchez MA (2022) Characterization of the UHI in Zaragoza (Spain) using a quality-controlled hourly sensor-based urban climate network. Urban Climate 44:101207. 10.1016/j.uclim.2022.101207

Beck C, Straub A, Breitner S, Cyrys J, Philipp A, Rathmann J, Schneider A, Wolf K, Jacobeit J (2018a) Air temperature characteristics of local climate zones in the Augsburg urban area (Bavaria, southern Germany) under varying synoptic conditions. Urban Clim 25(152):166. 10.1016/j.uclim.2018.04.007

Beck HE, Zimmermann NE, McVicar TR, Vergopolan N, Berg A, Wood EF (2018b) Present and future Köppen-Geiger climate classification maps at 1-km resolution. Sci Data 5:180214. 10.1038/sdata.2018.214 PubMed PMC

Bischl B, Richter J, Becker M, Binder M, Lang M, Pielok T, Coors S, Thomas J, Ullmann T, Boulesteix A-L, Deng D, Lindauer M (2023) Hyperparameter optimization: Foundations, algorithms, best practices, and open challenges. Wires Data Mining Knowl Discov 13:1484. 10.1002/widm.1484

Breiman L (2001) Random forests. Mach Learn 45:5–32. 10.1023/A:1010933404324

Briesch M, Sobania D, Rothlauf F (2022) The randomness of input data spaces is an a priori predictor for generalization. In: Bergmann R, Malburg L, Rodermund SC, Timm IJ (eds) KI 2022: advances in artificial intelligence. Springer International Publishing, Cham, pp 17–30

Cedro A, Nowak G (2006) Effects of climatic conditions on annual tree ring growth of the Platanus × hispanica ‘Acerifolia’ under urban conditions of Szczecin. Dendrobiology 55:11–17

Chen T, Guestrin C (2016) XGBoost: a scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. ACM, San Francisco California USA. pp 785–794

Corell M, Girón IF, Galindo A, Torrecillas A, Torres-Sánchez R, Pérez-Pastor A, Moreno F, Moriana A (2014) Using band dendrometers in irrigation scheduling: influence of the location inside the tree and comparison with point dendrometer. Agric Water Manag 142:29–37. 10.1016/j.agwat.2014.04.005

Deutscher Wetterdienst (2024) Mainz-Lerchenberg Station data. https://www.dwd.de/DE/leistungen/klimadatendeutschland/vielj_mittelwerte.html. Accessed 4 Apr 2024

Dormann CF, Elith J, Bacher S, Buchmann C, Carl G, Carré G, Marquéz JRG, Gruber B, Lafourcade B, Leitão PJ, Münkemüller T, McClean C, Osborne PE, Reineking B, Schröder B, Skidmore AK, Zurell D, Lautenbach S (2013) Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36:027–046. 10.1111/j.1600-0587.2012.07348.x

Dulamsuren C, Coners H, Leuschner C, Hauck M (2023) Climatic control of high-resolution stem radius changes in a drought-limited southern boreal forest. Trees 37:797–810. 10.1007/s00468-022-02384-z

Eischeid JK, Baker CB, Karl TR, Diaz HF (1995) The quality control of long-term climatological data using objective data analysis. J Appl Meteorol 34(12):2787–2795. 10.1175/1520-0450(1995)034%3c2787:TQCOLT%3e2.0.CO;2

Fatima S, Hussain A, Amir SB, Ahmed SH, Aslam SMH (2023) XGBoost and random forest algorithms: an in depth analysis. Pak J Sci Res 3(1):26–31. 10.57041/pjosr.v3i1.946

Fix E, Hodges JL (1951) Discriminatory analysis. Nonparametric discrimination consistency properties. USA Air Force School of Aviation Medicine, Randolph Field

Friedman JH (2001) Greedy function approximation: a gradient boosting machine. Ann Statist 29(5):1189–1232. 10.1214/aos/1013203451

Geiger RS, Cope D, Ip J, Lotosh M, Shah A, Weng J, Tang R (2021) “Garbage in, garbage out” revisited: what do machine learning application papers report about human-labeled training data? Quant Sci Stud. 10.1162/qss_a_00144

Géron A (2019) Hands-on machine learning with Scikit-Learn, Keras & TensorFlow. Concepts, tools, and techniques to build intelligent systems, 2nd edn. O’Reilly, Sebastopol

Google Satellite Image (2024) Google Maps. In: AeroWest Airbus, CNES/Airbus, GeoBasis-DE/BKH, GeoContent, Landsat/Copernicus, Maxar Technologies, Map data: GeoBasis-DE/BKG, Google

Haeni M, Knüsel S, Wilhelm M, Peters RL, Zweifel R (2020) treenetproc—Clean, process and visualise dendrometer data. R package

Head T, MechCoder GL, Shcherbatyi I (2018) scikit-optimize/scikit-optimize: v0. 5.2. Version v0 5

Hoerl AE, Kennard RW (2000) Ridge regression: biased estimation for nonorthogonal problems. Technometrics 42(1):80–86. 10.1080/00401706.2000.10485983

Iakovoglou V, Thompson J, Burras L, Kipper R (2001) Factors related to tree growth across urban-rural gradients in the Midwest, USA. Urban Ecosyst 5:71–85. 10.1023/A:1021829702654

King G, Fonti P, Nievergelt D, Büntgen U, Frank D (2013) Climatic drivers of hourly to yearly tree radius variations along a 6 °C natural warming gradient. Agric Meteorol 168:36–46. 10.1016/j.agrformet.2012.08.002

Knüsel S, Peters RL, Haeni M, Wilhelm M, Zweifel R (2021) Processing and extraction of seasonal tree physiological parameters from stem radius time series. Forests 12:765. 10.3390/f12060765

Landeshauptstadt Mainz (2024) Jahresbericht Juli 2022–Juni 2023: Hauptamt Abteilung Öffentlichkeitsarbeit und Protokoll sowie städtische Ämter, Mainz

Lindén J, Fonti P, Esper J (2016) Temporal variations in microclimate cooling induced by urban trees in Mainz, Germany. Urban Urban Green 20:198–209. 10.1016/j.ufug.2016.09.001

Lukovic M, Zweifel R, Thiry G, Zhang C, Schubert M (2022) Reconstructing radial stem size changes of trees with machine learning. J R Soc Interface 19:20220349. 10.1098/rsif.2022.0349 PubMed PMC

Lv H, Dermann A, Dermann F, Petridis Z, Köhler M, Saha S (2024) Comparable diameter resulted in larger leaf area and denser foliage in the park trees than in street trees: a study on Norway maples of Karlsruhe city, Germany. Heliyon 10:e23647. 10.1016/j.heliyon.2023.e23647 PubMed PMC

Moser-Reischl A, Rahman MA, Pauleit S, Pretsch H, Rötzer T (2019) Growth patterns and effects of urban micro-climate on two physiologically contrasting urban tree species. Landsc Urban Plan 183:88–99. 10.1016/j.landurbplan.2018.11.004

Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, Blondel M, Prettenhofer P, Weiss R, Dubourg V, Vanderplas J, Passos A, Cournapeau D, Brucher M, Perrot M, Duchesnay E (2011) Scikit-learn: machine learning in python. J Mach Learn Res 12:2825–2830. 10.48550/arXiv.1201.0490

R Core Team (2021) R: a language and environment for statistical computing. Austria, Vienna

Rainio O, Teuho J, Klén R (2024) Evaluation metrics and statistical tests for machine learning. Sci Rep 14:6086. 10.1038/s41598-024-56706-x PubMed PMC

Salomón RL, Peters RL, Zweifel R, Sass-Klaassen UGW, Stegehuis AI, Smiljanic M, Poyatos R, Babst F, Cienciala E, Fonti P, Lerink BJW, Lindner M, Martinez-Vilalta J, Mencuccini M, Nabuurs G-J, van der Maaten E, von Arx G, Bär A, Akhmetzyanov L, Balanzategui D, Bellan M, Bendix J, Berveiller D, Blaženec M, Čada V, Carraro V, Cecchini S, Chan T, Conedera M, Delpierre N, Delzon S, Ditmarová Ľ, Dolezal J, Dufrêne E, Edvardsson J, Ehekircher S, Forner A, Frouz J, Ganthaler A, Gryc V, Güney A, Heinrich I, Hentschel R, Janda P, Ježík M, Kahle H-P, Knüsel S, Krejza J, Kuberski Ł, Kučera J, Lebourgeois F, Mikoláš M, Matula R, Mayr S, Oberhuber W, Obojes N, Osborne B, Paljakka T, Plichta R, Rabbel I, Rathgeber CBK, Salmon Y, Saunders M, Scharnweber T, Sitková Z, Stangler DF, Stereńczak K, Stojanović M, Střelcová K, Světlík J, Svoboda M, Tobin B, Trotsiuk V, Urban J, Valladares F, Vavrčík H, Vejpustková M, Walthert L, Wilmking M, Zin E, Zou J, Steppe K (2022) The 2018 European heatwave led to stem dehydration but not to consistent growth reductions in forests. Nat Commun 13(1):28. 10.1038/s41467-021-27579-9 PubMed PMC

Schwarz L, Sobania D, Rothlauf F (2024) On relevant features for the recurrence prediction of urothelial carcinoma of the bladder. Int J of Med Inform 186:105414. 10.1016/j.ijmedinf.2024.105414 PubMed

Ushey K, Allaire JJ, Tang Y (2024) Reticulate: interface to “python”. R package version 1.39.0. https://github.com/rstudio/reticulate, https://rstudio.github.io/reticulate/

Ying X (2019) An overview of overfitting and its solutions. J Phys Conf Ser 1168:022022. 10.1088/1742-6596/1168/2/022022

Zhang Y, Gao L, Deng Y, Huang Q, Yuan Y, Shi X (2024) Seasonal aridity regulates drivers and temporal variability of wood phenology: a meta-analysis of dendrometer monitoring data across the northern hemisphere. Dendrochronologia 85:126201. 10.1016/j.dendro.2024.126201

Ziaco E, Biondi F (2018) Stem circadian phenology of four pine species in naturally contrasting climates from sky-island forests of the western USA. Forests 9(7):396. 10.3390/f9070396

Zweifel R, Haeni M, Buchmann N, Eugster W (2016) Are trees able to grow in periods of stem shrinkage? New Phytol 211:839–849. 10.1111/nph.13995 PubMed