Font Size: a A A

Spatio-Temporal Characteristics And Trends Of Forest Fire Points In Jiangxi Province

Posted on:2020-05-30Degree:MasterType:Thesis
Country:ChinaCandidate:X L GuFull Text:PDF
GTID:2393330575960385Subject:Physical geography
Abstract/Summary:PDF Full Text Request
Forest fire is one of the most important disturbances in forest ecosystems,which significantly alters forest landscape structure and function worldwide.Jiangxi Province is one of the typical distribution areas of subtropical evergreen broad-leaved forests in China.The frequency of forest fires in this region is high,which has a significant impact on social economy and sustainable development of forest ecosystem.In the context of global climate change,identifying the spatial-temporal differentiation of forest fires in this region and revealing the impact of climate change on the spatial distribution pattern of forest fires are the basis for formulating scientific forest fire management policies,which are crucial to maintaining the health of forest ecosystem.It can provide scientific guidance for forest fire management and fire prevention resource allocation.Therefore,based on the data of MODIS fire products(MCD14ML)from 2001 to 2015 in Jiangxi province,this study used mathematical statistics,spatial analysis and information entropy theory to study the spatial and temporal distribution characteristics of forest fires in Jiangxi province from 2001 to 2015.Based on the data of climate,vegetation,topography and human activities,the relative importance and marginal effect of the factors influencing forest fire were analyzed by using the enhanced regression tree model.The GFDL-CM3 and GISS-E2-R climate change models were used to predict the forest fire distribution in Jiangxi province in 2050 and 2070 under three greenhouse gas emission scenarios(RCP2.6,RCP4.5 and RCP8.5),and the subject operating characteristics(ROC curve)and confusion matrix were used to evaluate the prediction accuracy of the model.The results show that:(1)the interannual fluctuation of forest fires in Jiangxi province was large from 2001 to 2015,with two peaks in 2004 and 2008.Seasonal changes are obvious,mainly in winter and spring;(2)there are three significant spatial hot spots for forest fire occurrence,which are located in the southwest of Ganzhou city,the northwest of Ji 'an city and the south of Fuzhou city respectively;(3)the spatial and temporal comprehensive distribution of forest fire is an aggregation model with a high degree of aggregation;(4)the annual temperature and altitude showed strong correlations with the occurrence of forest fire in Jiangxi province,and the annual precipitation,distance to residential areas,population density,and distance to roads had weaker correlations with the occurrence of forest fires;(5)the AUC values(area values under the ROC curve)of the training data(70%)and the verification data(30%)were both 0.736,and the accuracy of the obfuscations matrix in predicting the fire point was 67.8%,indicating that the model could well predict the occurrence of forest fires in the study area.;(6)the increasing in forest fire occurrence was highest under the future climate scenario of RCP8.5;(7)under the future climate scenario,the changes of forest fires in each region in the future are different.Compared with the distribution of forest fires in 2050 and 2070 under the current climate condition,the increases of ganzhou city and yingtan city are relatively obvious,while the changes of other regions are not obvious.This study demonstrates the occurrence of forest fires on regional scale is not completely random,contrary to present significant temporal clustering characteristics,and under the background of climate change in the future,the occurrence of forest fire has a tendency to increase,therefore,in jiangxi province,the forestry administration departments to strengthen the high incidence and potential occurring fires the forest monitoring and management,and increase the intensity of fire prevention propaganda,enhance people's awareness of forest fire prevention.
Keywords/Search Tags:forest fire, hotspot analysis, MCD14ML, temporal and spatial features, climate change, boosted regression tree
PDF Full Text Request
Related items