Collaborative Research Team Project #04
Evolving Marked Point Processes with Application to Wildland Fire Regime Modeling
This project explores statistical methods for quantifying and mapping fire risk, with a focus on the wildland-urban interface.
Research Category: Ecology & Environment
Why Study Wildland Fires?
Our environment is evolving. High-profile wildfires have heightened awareness and concern surrounding wildland fires on populated landscapes. It is important to understand how biotic feedback, climate, the weather, and fire suppression interact to impact fire risk.
Canadian forest fire management agencies keep detailed historical records on fires, weather and fire suppression. These large and complex spatio-temporal data sets provide a rich set of information for investigation.
Spatially and temporally-explicit methods are needed that quantify and map the risk of:
- large fires,
- spread events,
- variation in fire severity,
- local concentrations of fires,
- and the potential for a suppressed fire to escape initial attack.
Through the development and application of sophisticated statistical methodology, we can produce future fire regimes that are data-driven and effective.
Areas of Exploration
Marked Point Processes
Includes using marked point process (MPP) methods to model fire ignition points in historical data from Alberta and other provinces. This takes into account weather conditions and features of the natural and built environments.
Interacting Particle Systems
Includes investigating how best to model smouldering and fire spread using underlying interacting particle systems (IPS).
Fire Regime Models
Includes developing visualization tools for fire managers and property insurers. This is an important outcome of the collaboration with provincial ministries and with the Institute for Catastrophic Loss Reduction (a non-profit institute affiliated with the University of Western Ontario).
Solving Global Challenges
Research Team’s Goal
To develop statistical methods for quantifying and mapping fire risk. Specifically, to model MPP data aggregated over moderate to large temporal and spatial scales.
People Behind the Project
John Braun | University of British Columbia-Okanagan
Douglas Woolford | University of Western Ontario
Patrick Brown | University of Toronto
David Martell | University of Toronto
Jamie Stafford | University of Toronto
Charmaine Dean | University of Western Ontario
Bruce Jones | University of Western Ontario
Steve Cumming | Université Laval
Thierry Duchesne | Université Laval
Mike Flannigan | University of Alberta
Joan Hu | Simon Fraser University
Michael Wotton | Canadian Forest Service
- W.J. Braun and J.E. Stafford. Estimating Conditional and Multivariate Density Functions in the Presence of Interval Censoring. Submitted to Environmetrics, August, 2015 and currently under revision.
- W.J. Braun, X.J. Hu, and X. Kang. Data Sharpening Guided by Global Constraint in Local Regression. Submitted to Computational Statistics and Data Analysis, October, 2015.
- Ainsworth, L.M. and Dean, C.B. (2016), Zero-inflated Spatial Models. Advances and Challenges in Parametric and Semi-parametric Analysis for Correlated Data, Lecture Notes in Statistics- Proceedings of the 2015 International Symposium in Statistics, 75-96, doi: 10.1007/978-3-319-31260-6.
- Hosseini, R., Newlands, N.K., Dean, C.B., Takemura, A. (2015), Statistical Modeling of Soil Moisture, Integrating Satellite Remote-Sensing (SAR) and Ground-Based Data, Remote Sensing 7, 2752-2780.
- Wolters, M.A., and Dean, C.B. (2015), Issues in the Identification of Smoke in Hyperspectral Satellite Imagery: a Machine Learning Approach, Current Air Quality Issues, Mejadkoorki, F. (ed.), In-Tech, doi: 10.5772/60214.
- Marchal, Jean, Steve G. Cumming, and Eliot JB McIntire. “Exploiting Poisson additivity to predict fire frequency from maps of fire weather and land cover in boreal forests of Québec, Canada.” Ecography (2016).
- Da Zhong (Dexen) Xi, Charmaine Dean, Steve Taylor; Joint Models for the Duration and Size of BC Forest Fires (paper in progress).
- Alisha Albert-Green, W. John Braun, Charmaine Dean; A Spatio-Temporal Cluster Process for Modelling Storm Cells (Paper in progress)
- Alisha Albert-Green; Thesis: Joint Models for Spatial and Spatio-Temporal Point Processes.
- Peter Hall, John Braun, Thierry Duchesne (2017). Empirical Fourier methods for interval censored data. (paper in progress).
- Pier-Olivier Tremblay, Steve Cumming, Thierry Duchesne (2017). Finding factors associated with fire size using a two-stage survival analysis approach. (paper in progress).
- Amy A. Morin, Alisha Albert-Green, Douglas G. Woolford and David L. Martell. A framework for exploring spatial differences in the control time of forest fires in Ontario, Canada. (paper in progress).
- Han, L., Braun, W.J. and Loeppky, J. Random coefficient minification models. Statistical Papers, 1–22, 2018.
- Xiong, Y., Hu, X.J. and Braun, W.J. On the use of Moran’s I statistic with spatio-temporal observations. Submitted to Computational Statistics and Data Analysis, 2016.
- Marchal, J., Cumming, S.G., and McIntire, E.J.B. (01/2017) Exploiting Poisson additivity to predict fire frequency from maps of fire weather and land cover in boreal forests of Québec, Canada. Ecography 40(1), 200-209. 10.1111/ecog.01849.
- Marchal, J., Cumming, S.G., and McIntire, E.J.B. (01/2017) Land cover, more than monthly fire weather, drives fire-size distribution in Southern Québec forests: Implications for fire risk management.PLoS ONE 12(6), e0179294. https://doi.org/10.1371/journal.pone.0179294
- Tremblay, P-O., Duchesne, T., and Cumming, S. (2018) Survival analysis and classification methods for forest fire size. Accepted for publication in PLOS ONE
- Albert-Green, A, Braun, W. J., Dean, C.B. and Miller, C. A Hierarchical Point Process with Application to Storm Cell Modelling. Under revision for Canadian Journal of Statistics.
- Albert-Green, A., Dean, C.B. and Braun, W.J. A General Framework for the Joint Modelling of Multivariate Zero-Inflated Spatial Processes. Submitted to the Journal of Agricultural, Biological and Environmental Statistics.
- Xiong, Hu and Braun. On the use of Moran’s I statistic with spatio-temporal observations. Submitted to the Journal of Nonparametric Statistics.
- Wang, Braun and Woolford. Fitting a stochastic fire spread model to data. To be submitted to Environmetrics.
- Wang, Braun and Brown. Data sharpening to reduce bias in local polynomial derivative estimation (in preparation).
- Albert-Green, A., Guttorp, P. and Thorarinsdottir, T. Does Bayes Beat Squinting? Estimating Unobserved Aspects of a Spatial Cluster Process (in preparation)
Explore More Stories
Find Related Programs
Evolving Marked Point Processes with Application to Wildland Fire Regime Modeling is a Collaborative Research Team project. This program tackles complex problems through a three-year research and training agenda.
CANSSI offers approximately $200,000 for this type of project, which requires a team of faculty, postdocs, and students.