Forest Fire Susceptibility Modeling Using a Convolutional Neural Network for Yunnan Province of China

Abstract

Forest fires have caused considerable losses to ecologies, societies, and economies worldwide. To minimize these losses and reduce forest fires, modeling and predicting the occurrence of forest fires are meaningful because they can support forest fire prevention and management. In recent years, the convolutional neural network (CNN) has become an important state-of-the-art deep learning algorithm, and its implementation has enriched many fields. Therefore, we proposed a spatial prediction model for forest fire susceptibility using a CNN. Past forest fire locations in Yunnan Province, China, from 2002 to 2010, and a set of 14 forest fire influencing factors were mapped using a geographic information system. Oversampling was applied to eliminate the class imbalance, and proportional stratified sampling was used to construct the training/validation sample libraries. A CNN architecture that is suitable for the prediction of forest fire susceptibility was designed and hyperparameters were optimized to improve the prediction accuracy. Then, the test dataset was fed into the trained model to construct the spatial prediction map of forest fire susceptibility in Yunnan Province. Finally, the prediction performance of the proposed model was assessed using several statistical measures-Wilcoxon signed-rank test, receiver operating characteristic curve, and area under the curve (AUC). The results confirmed the higher accuracy of the proposed CNN model (AUC 0.86) than those of the random forests, support vector machine, multilayer perceptron neural network, and kernel logistic regression benchmark classifiers. The CNN has stronger fitting and classification abilities and can make full use of neighborhood information, which is a promising alternative for the spatial prediction of forest fire susceptibility. This research extends the application of CNN to the prediction of forest fire susceptibility.

Comparisons of Diverse Machine Learning Approaches for Wildfire Susceptibility Mapping

“…Remote sensing and the geographic information system (GIS) are significant tools for wildfire analysis and susceptibility mapping. Natural hazards like landslides, floods, earthquakes, and wildfires have been assessed using diverse methodologies like analytical hierarchical process (AHP) [12], frequency ratio (FR) [13,14], support vector machines (SVM) [15], analytical network process (ANP) [16], random forest (RF) [17], Dempster-Shafer [14,18,19], artificial neural networks (ANN) [20], decision tree (DT) [21], logistic regression (LR) [22][23][24] convolution neural network (CNN) [20,25] and evidence belief function (EBF) [26]. The results of some of these methodologies and models were evaluated and compared to each other.…”

A review of machine learning applications in wildfire science and management

“…Bashari et al (2016) noted that BNs may be useful because they allow probabilities to be updated when new observations become available. SVM was also used by a number of authors as a benchmark for other ML methods (Ghorbanzadeh et al 2019b;Gigović et al 2019;Hong et al 2018;Jaafari and Pourghasemi 2019;Thach et al 2018;Rodrigues and De la Riva 2014;Sachdeva et al 2018;Tehrany et al 2018;Bui et al 2017;van Breugel et al 2016;Zhang et al 2019), but as we discuss later, it did not perform as well as other methods with which it was being compared.…”

“…In particular, for fire detection, compared CNNs with SVM and found that CNNs were more accurate, while Zhao et al (2018) similarly found CNNs superior to SVMs and ANNs. For firesusceptibility mapping, Zhang et al (2019) found CNNs to be more accurate than RF, SVMs, and ANNs. For time series forecasting problems, Liang et al (2019) found that LSTMs outperformed ANNs.…”

“…Bui et al (2017) used PSO to choose inputs to a neural fuzzy model and found that this improved results. Zhang et al (2019) also considered the information gain ratio for feature selection. As noted in Moritz et al (2012) and Mayr et al (2018), one should also take spatial autocorrelation into account when modeling fire probabil-ities spatially.…”

Maximum entropy method-based forest fire prediction mapping of Sikkim Himalaya.

“…Another important limitation of machine learning method, to be speci c methods involving arti cial neural networking, is that, they achieve e ciency and accuracy at the cost of interpretability of the model(Nami et al 2018;Tien Bui, Le, and Hoang 2018;Ghorbanzadeh, Kamran, and Blaschke 2019;Tehrany et al 2019;Tien Bui, Hoang, and Samui 2019;Zhang, Wang, and Liu 2019).…”