Statistical modeling of tree failures during storms

Abstract The failure of trees during storms imposes strong economic and societal costs. Statistical modeling for predicting the probability of a tree failing during storms has the potential to help improve tree risk management. The purpose of this study is to explore the potential predictability of tree failure using advanced predictive modeling approach. These models also have broader applicability for modeling failures of technical systems during adverse weather events. To train and test models, we use a data set from a real case study in Massachusetts, USA. We compare the out-of-sample predictive accuracy of several machine learning models including logistic regression, classification and regression trees, multivariate adaptive regression splines, artificial neural network, naive-Bayes regression, random forest, boosting, and an ensemble model of boosting and random forest. Our results demonstrate that the ensemble model of boosting and random forest achieves the best prediction accuracy in predicting the failure probability of trees for the case study storm. Our results can help tree care professionals make better decisions to reduce the risk of tree failure prior to the storm.

[1]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1997, EuroCOLT.

[2]  Julio J. Melero,et al.  Data-driven learning framework for associating weather conditions and wind turbine failures , 2018, Reliab. Eng. Syst. Saf..

[3]  E. Thomas Smiley,et al.  Impact of assessor on tree risk assessment ratings and prescribed mitigation measures , 2017 .

[4]  B. Gardiner,et al.  Management of forests to reduce the risk of abiotic damage — a review with particular reference to the effects of strong winds , 2000 .

[5]  R. O’Brien,et al.  A Caution Regarding Rules of Thumb for Variance Inflation Factors , 2007 .

[6]  Chao Hu,et al.  Ensemble of data-driven prognostic algorithms for robust prediction of remaining useful life , 2011, 2011 IEEE Conference on Prognostics and Health Management.

[7]  Ali Mosleh,et al.  Methodology for the use of experimental data to enhance model output uncertainty assessment in thermal hydraulics codes , 2010, Reliab. Eng. Syst. Saf..

[8]  Brian Kane.  Determining parameters related to the likelihood of failure of red oak (Quercus rubra L.) from winching tests , 2014, Trees.

[9]  Edward N. Rappaport,et al.  Fatalities in the United States from Atlantic Tropical Cyclones: New Data and Interpretation , 2014 .

[10]  Alexis Achim,et al.  Wind loading of trees: influence of tree size and competition , 2010, European Journal of Forest Research.

[11]  Jonathan M. Garibaldi,et al.  A 'non-parametric' version of the naive Bayes classifier , 2011, Knowl. Based Syst..

[12]  Wei Wang,et al.  Non-linear partial least squares response surface method for structural reliability analysis , 2017, Reliab. Eng. Syst. Saf..

[13]  Joaquim F. Silva,et al.  Finding occupational accident patterns in the extractive industry using a systematic data mining approach , 2012, Reliab. Eng. Syst. Saf..

[14]  Gregory A. Dahle,et al.  A review of factors that affect the static load- bearing capacity of urban trees , 2017 .

[15]  Gregory A. Dahle,et al.  Tree Biomechanics Literature Review: Dynamics , 2014, Arboriculture & Urban Forestry.

[16]  Dominique Guyon,et al.  Mechanistic and statistical approaches to predicting wind damage to individual maritime pine (Pinus pinaster) trees in forests , 2016 .

[17]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[18]  T. M. Logan,et al.  Predictive models in horticulture: A case study with Royal Gala apples , 2016 .

[19]  J. Chambers,et al.  Mechanical vulnerability and resistance to snapping and uprooting for Central Amazon tree species , 2016 .

[20]  J. Faraway.  Extending the Linear Model with R: Generalized Linear, Mixed Effects and Nonparametric Regression Models , 2005 .

[21]  Sanjay R. Arwade,et al.  Analysis of the probability of failure for open-grown trees during wind storms , 2014 .

[22]  Ivo Paixao de Medeiros,et al.  Remaining useful life estimation in aeronautics: Combining data-driven and Kalman filtering , 2018, Reliab. Eng. Syst. Saf..

[23]  Andy Liaw,et al.  Classification and Regression by randomForest , 2007 .

[24]  Marcello Braglia,et al.  The classification and regression tree approach to pump failure rate analysis , 2003, Reliab. Eng. Syst. Saf..

[25]  J. Dwyer,et al.  Understanding the Benefits and Costs of Urban Forest Ecosystems , 2007 .

[26]  Chris J. Peterson,et al.  Consistent influence of tree diameter and species on damage in nine eastern North America tornado blowdowns , 2007 .

[27]  Ali Mosleh,et al.  Structured treatment of model uncertainty in complex thermal-hydraulics codes: Technical challenges, prospective and characterization , 2011 .

[28]  Seth D. Guikema,et al.  Hybrid data mining-regression for infrastructure risk assessment based on zero-inflated data , 2012, Reliab. Eng. Syst. Saf..

[29]  H. Peltola,et al.  A mechanistic model for assessing the risk of wind and snow damage to single trees and stands of Scots pine, Norway spruce, and birch , 1999 .

[30]  William N. Venables,et al.  Modern Applied Statistics with S , 2010 .

[31]  Brian Kane,et al.  The Effects of Pruning on Drag and Bending Moment of Shade Trees , 2008, Arboriculture & Urban Forestry.

[32]  Thomas Lengauer,et al.  ROCR: visualizing classifier performance in R , 2005, Bioinform..

[33]  Johan Östberg,et al.  Tree inventories in the urban environment , 2013 .

[34]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[35]  Raha Akhavan-Tabatabaei,et al.  An ensemble classifier to predict track geometry degradation , 2017, Reliab. Eng. Syst. Saf..

[36]  Barry Gardiner,et al.  Observations and predictions of wind damage to Larix kaempferi trees following thinning at an early growth stage , 2017 .

[37]  B. Courbaud,et al.  Development of an individual tree-based mechanical model to predict wind damage within forest stands , 2004 .

[38]  David L. Woodruff,et al.  Generating short‐term probabilistic wind power scenarios via nonparametric forecast error density estimators , 2017 .

[39]  Brian Kane,et al.  Tree failure following a windstorm in Brewster, Massachusetts, USA , 2008 .

[40]  Yoav Freund,et al.  Experiments with a New Boosting Algorithm , 1996, ICML.

[41]  David Rines,et al.  Measuring urban forestry performance and demographic associations in Massachusetts, USA , 2011 .

[42]  Thomas W. Schmidlin,et al.  Human fatalities from wind-related tree failures in the United States, 1995–2007 , 2009 .

[43]  Wei-Yin Loh,et al.  Classification and regression trees , 2011, WIREs Data Mining Knowl. Discov..

[44]  Daria Battini,et al.  Estimating reliability characteristics in the presence of censored data: A case study in a light commercial vehicle manufacturing system , 2010, Reliab. Eng. Syst. Saf..

[45]  Srikanta Mishra,et al.  Application of classification trees in the sensitivity analysis of probabilistic model results , 2003, Reliab. Eng. Syst. Saf..

[46]  Barry Gardiner,et al.  A review of mechanistic modelling of wind damage risk to forests , 2008 .

[47]  Daniel G. Morrow,et al.  Using neural networks to assess flight deck human-automation interaction , 2013, Reliab. Eng. Syst. Saf..