Spatial pattern prediction of forest wildfire susceptibility in Central Yunnan Province, China based on multivariate data

Abstract

Wild res are an important disturbance factor in forest ecosystems. Assessing the probability of forest wild res can assist in forest wild re prevention, control, and supervision. The logistic regression model is widely used to forecast the probability, spatial patterns, and drivers of forest wild res. This study used logistic regression to establish a spatial prediction model for forest wild re susceptibility, which was applied to evaluate the risk of forest wild res in Central Yunnan Province (CYP), China. A forest wild re risk classi cation was implemented for CYP using forest burn scar data for 2001 to 2020 and the logistic spatial prediction model for forest wild re susceptibility. Climate, vegetation, topographical, human activities, and location were selected as forest wild re prediction variables. The results showed that: (1) The distributions of temperature, vegetation coverage, distance to water bodies, distance to roads, and precipitation were positively correlated with the occurrence of forest wild res. Elevation, relative humidity, the global vegetation moisture index, wind speed, slope, latitude, and distance to residential areas were negatively correlated with the occurrence of forest wild res. (2) The results of the logistic spatial prediction model for forest wild re susceptibility showed a good t to observed data, with an overall simulation probability of 81.6%. The optimal threshold for spatial prediction for forest wild re susceptibility in CYP was determined to be 0.414. A signi cance level of a selected model variable of < 0.05 resulted in an area under the receiver operating characteristic curve (AUC) of 0.882-0.890. (3) Forest wild re prevention efforts should focus on Southwest Yuxi City and southern Qujing City since they accounted for a high proportion of the areas at high risk of forest wild res. Other localities should adjust forest wild re prevention measures according to local conditions and strengthen existing wild re prevention and emergency resource planning and allocation. (4) Some factors contributing to forest wild res where different among the different areas. Forest wild re risk factors had different degrees of impact under different spatial and temporal scales. The spatial relationships between wild re disasters and in uencing factors should be established in areas with heterogeneous environmental conditions for the selection of relevant factors.

Assessment of Wildfire Susceptibility and Wildfire Threats to Ecological Environment and Urban Development Based on GIS and Multi-Source Data: A Case Study of Guilin, China

“…The distance from the water system determines human activity and vegetation growth status [44]. FVC indicates the density of surface vegetation, and a high coverage of vegetation signifies more fuel, thereby increasing the likelihood of wildfire ignition and spread [21]. The calculation formula is as follows:…”

“…In recent years, various ML models such as logistic regression (LR) [16], artificial neural networks (ANN) [17], support vector machines (SVM) [18], random forests (RF) [19], and gradient boosting decision trees (GBDT) [20] have been widely applied in the evaluation of wildfire susceptibility. They establish a connection between wildfire data and different conditioning factors, and can better fit data samples and highlight the nonlinear relationship between wildfires and factors, thus achieving more accurate prediction results [21].…”

Geospatial XAI: A Review

“…More than half of the research had the overall goal of prediction in nature-related use cases. This includes predicting slope failures/landslides [13,15,19,23] or avalanche hazards [20]; mapping natural disasters such as earthquakes [12,27] and wildfires [16,17]; or modeling other relationships with temperature [22,24,26] or pollution [18,21]. The next two major categories are research in traffic and transport and research in human-related use cases.…”

“…There are two ways of calculating feature importance. One would be to use linear models and take the regression coefficients [17]. The other one is to use permutation [16,21,37].…”