Prediction model of water content in surface dead fuel based on convolution neural network and meteoro-logical factors regression.

Water content of surface fuels is an important indicator of forest fire risk level and fire behavior, and the prediction model of which has a significant effect on fire risk prediction and management. Based on field meteorological factors of Quercus mongolica and Pinus sylvestris var. mongolica forests and water content data of dead fuels on the ground, we conducted the relative importance ranking of meteorological factors random forest and Pearson correlation analysis, and predicted water content of fuels using deeply learned convolutional neural network and meteorological factors regression. The results showed that water content of Q. mongolica fuels in the wild was significantly higher than that of P. sylvestris var. mongolica. The results of random forest showed that the factors that had significant effect on water content of fuel were humidity, temperature, rainfall, wind speed, and solar radiation, with the importance ranking from the largest to the smallest. Results of correlation analysis showed that temperature, humidity, and rainfall of current day had a significant impact on water content of fuels, and certain correlations were observed between meteorological factors. The prediction R2 of the convolutional neural network model for the surface fuel water content of Q. mongolica and P. sylvestris var. mongolica forest was 0.928 and 0.905, the mean absolute error (MAE) was 6.1% and 8.1%, and the mean relative error (MRE) was 8.9% and 4.2%, respectively. However, the R2, MAE, MRE of meteorological factors regression were 0.495 and 0.525, 30.5% and 39.5%, 52.1% and 32.6%, respectively. The precision of convolution neural network model was significantly higher than that of meteorological factors regression. Our results showed that the deeply learned convolutional neural network could provide some reference for the prediction of fuel water content in the future, and effectively support higher level forest fire management.

Prediction models of fire spread rate of Pinus koraiensis plantation's surface fuel.

Pinus koraiensis plantation has high fire risks due to high oil content in branches and leaves. The spread of surface fire is the main way of forest fire expansion. Understanding the surface fire spread rate can provide scientific guidance for fire fighting. We carried out a laboratory experiment with surface fuel of Pinus koraiensis plantation in Maoershan area of Heilongjiang Province. We set different levels of fuel moisture contents (5%, 15%, 25%), fuel loads (0.5, 0.7, 0.9, 1.1 kg·m-2), and slope (0°, 10°, 20°, 30°, 40°) to simulate the characteristics of fuel bed in the field, and quantified the spread rate by thermocouple method. We further compared and analyzed the prediction accuracy of Rothermel model, modified Rothermel model and random forest model, and evaluated the optimal model for predicting the surface fire spread rate of P. koraiensis plantation. The results showed that the overall efficacy of directly using the Rothermel model to predict the surface fire spread rate of P. koraiensis plantations was good, but the prediction result of the spread rate under the conditions of high slope and high moisture content was not satisfied. The Rothermel model after refitting the slope parameters and the random forest model had good prediction efficacy and similar prediction accuracy. The random forest model needed to be further evaluated and verified due to its own characteristics. The modified Rothermel model was more suitable for predicting the surface fire spread rate of P. koraiensis plantations at a slope range of 0°-40° than the others.

[Pyrolysis and gas emissions characteristics of six tree species in Heilongjiang Province, China]. / 黑龙江省6种乔木叶片热解特性及气体释放特征.

Forest fuels are the basis of fire occurrences, while ground dead fuels are an important part of forest fuels. Undestanding the pyrolysis characteristics and gas emissions of forest fuels is of great significance to explore the effects of forest fire on atmospheric environment and carbon balance, as well as to prevent and combat forest fire. In this study, the thermogravimetric analysis and gas emission analysis were conducted on leaf litter of six tree species (Pinus sylvestris var. mongolica, Picea koraiensis, Fraxinus mandshurica, Juglans mandshurica, Quercus mongolica, Betula platyphylla) in Heilongjiang Province to explore the pyrolysis process and combustibility of forest fuels, to analyze their pyrolysis characteristics, pyrolysis kinetics characteristics, gas emission characteristics. A four-dimensional evaluation of their combustibility was conducted based on pyrolysis parameters. The results showed that the pyrolysis temperature of holocellulose in the leaves of those six tree species ranged in 143.31-180.48 ℃ at the beginning and 345.04-394.38 ℃ at the end, lignin pyrolysis temperature ranged in 345.04-394.38 ℃ at the beginning and 582.85-609.31 ℃ at the end. The pyrolysis of the six kinds of arbor blades during the pyrolysis process affected fuel ash content, quality and temperature of the total pyrolysis. The activation energies of two main pyrolysis stages of leaves of six tree species were 18.88-27.08 kJ·mol-1 and 13.25-27.54 kJ·mol-1, respectively, and the pre-exponential factors were 3.13-26.28 min-1 and 1.30-22.55 min-1. The holocellulose activation energy and pre-exponential factor of the pyrolysis stage for P. koraiensis, F. mandshurica, Q. mongolica, and B. platyphylla were greater than that of the lignin pyrolysis stage, while the opposite was true for P. sylvestris var. mongolica and J. mandshurica. The release amounts of CO and CO2 at the pyrolysis stage of the holocellulose was 535.16-880.11 mg·m-3 and 7004.97-10302.05 mg·m-3, and that at the pyrolysis stage of lignin was 240.31-1104.67 mg·m-3 and 20425.60-33946.68 mg·m-3, respectively. The release of CO and CO2 at the pyrolysis stage of healdellulose was less, but mass loss was greater than that at the pyrolysis stage of lignin. In the four-dimensional combustibility ranking of the six tree species leaves, B. platyphylla was the best ignitable, P. koraiensis was the most combustible, and P. sylvestris var. mongolica was the most sustainable and consumable. The ignitability was significantly positively correlated with pyrolysis kinetics parameters of the holocellulose, while the sustainability was negatively correlated with that of lignin.