TY - JOUR
T1 - Predicting wildfire burns from big geodata using deep learning
AU - Bergado, J.R.
AU - Persello, C.
AU - Reinke, Karin
AU - Stein, A.
PY - 2021/8
Y1 - 2021/8
N2 - Wildfire continues to be a major environmental problem in the world. To help land and fire management agencies manage and mitigate wildfire-related risks, we need to develop tools for mapping those risks. Big geodata—in the form of remotely sensed images, ground-based sensor observations, and topographical datasets—can help us characterize the dynamics of wildfire related events. In this study, we design a deep fully convolutional network, called AllConvNet, to produce daily maps of the probability of a wildfire burn over the next 7 days. We applied it to burns in Victoria, Australia for the period of 2006–2017. Fifteen factors that were extracted from six different datasets and resulted into 29 quantitative features, were selected as input to the network. We compared it with three baseline methods: SegNet, multilayer perceptron, and logistic regression. AllConvNet outperforms the other three baseline methods in four of the six quantitative metrics considered. AllConvNet and SegNet provide smoother and more regularized predicted maps, with SegNet providing greater sensitivity in dificriminating less wildfire-prone locations. Input feature statistical importance was measured for all the networks and compared against logistic regression coefficients. Total precipitation, lightning flash density, and land surface temperature occur to be consistently highly weighted by all models while terrain aspect components, wind direction components, certain land cover classes (such as crop field and woodland), and distance from power lines are ranked on the lower end. We conclude that wild-fire burn prediction methods based on deep learning present quantitative and qualitative gains.
AB - Wildfire continues to be a major environmental problem in the world. To help land and fire management agencies manage and mitigate wildfire-related risks, we need to develop tools for mapping those risks. Big geodata—in the form of remotely sensed images, ground-based sensor observations, and topographical datasets—can help us characterize the dynamics of wildfire related events. In this study, we design a deep fully convolutional network, called AllConvNet, to produce daily maps of the probability of a wildfire burn over the next 7 days. We applied it to burns in Victoria, Australia for the period of 2006–2017. Fifteen factors that were extracted from six different datasets and resulted into 29 quantitative features, were selected as input to the network. We compared it with three baseline methods: SegNet, multilayer perceptron, and logistic regression. AllConvNet outperforms the other three baseline methods in four of the six quantitative metrics considered. AllConvNet and SegNet provide smoother and more regularized predicted maps, with SegNet providing greater sensitivity in dificriminating less wildfire-prone locations. Input feature statistical importance was measured for all the networks and compared against logistic regression coefficients. Total precipitation, lightning flash density, and land surface temperature occur to be consistently highly weighted by all models while terrain aspect components, wind direction components, certain land cover classes (such as crop field and woodland), and distance from power lines are ranked on the lower end. We conclude that wild-fire burn prediction methods based on deep learning present quantitative and qualitative gains.
KW - Big geodata
KW - Convolutional networks
KW - Deep learning
KW - Input statistical importance
KW - Wildfire burn prediction
KW - ITC-ISI-JOURNAL-ARTICLE
KW - UT-Hybrid-D
UR - https://ezproxy2.utwente.nl/login?url=https://doi.org/10.1016/j.ssci.2021.105276
UR - https://ezproxy2.utwente.nl/login?url=https://library.itc.utwente.nl/login/2021/isi/bergado_pre.pdf
U2 - 10.1016/j.ssci.2021.105276
DO - 10.1016/j.ssci.2021.105276
M3 - Article
AN - SCOPUS:85104312573
VL - 140
JO - Safety science
JF - Safety science
SN - 0925-7535
M1 - 105276
ER -