Font Size: a A A

Analysis Of Spatiotemporal Variation And Vegetation Trend Of Burned Area From 1990 To 2019 In Southern Jiangxi

Posted on:2022-03-03Degree:MasterType:Thesis
Country:ChinaCandidate:X H ZhangFull Text:PDF
GTID:2493306494499244Subject:Cartography and Geographic Information System
Abstract/Summary:PDF Full Text Request
One of the main disturbance factors affecting the stability of forest ecosystems is fire.It’s rich of forest resources in southern Jiangxi,which is an important ecological barrier in Jiangxi Province,and plays a certain role in maintaining the balance of ecological system in southern Jiangxi Province.The study on the temporal and spatial variation of the burned area and the vegetation restoration trend in southern Jiangxi can not only provide scientific reference for management of forest fire and construction of forest fire prevention projects,but also have important practical significance for improving plan of fire prevention,maintaining the safety of ecological resources and the safety of people’s lives and property.Bases on the Google Earth Engine cloud platform and Landsat TM/ETM+/OLI image data from 1990 to 2019 in southern Jiangxi,this study makes layers of confuse source(water and terrain shadow),and uses the method of spectral mixture analysis to extract the information of burned area,and analyzes the characteristics of spatiotemporal variation for 30 years.Secondly,the effects of climate factors(annual mean temperature,annual precipitation,annual mean wind speed,relative humidity)and topographic factors(altitude,slope,slope aspect)on forest fires are analyzed by single factor analysis and Geodetector.Finally,Sen+Mann-Kendall method is used to analyze the variation trend of NDVI for 30 years in the study area,so as to characterize the trend of forest vegetation restoration and analyze the spatial distribution pattern of the trend change.The main conclusions are as follows:(1)Based on the Google Earth Engine cloud platform and spectral mixed analysis method,the average value of overall accuracy and Kappa coefficient are 88.38% and0.82,respectively,indicating a high overall accuracy.The results extracted in this study are basically consistent with the actual range and clear boundary of burned area from high-resolution image.The results of accuracy indicate that the spectral mixed analysis method is more reliable to extract the burned area at regional scale.(2)The temporal variation and spatial distribution of burned area in southern Jiangxi Province are significantly different from 1990 to 2019.The temporal variation of burned area show the alternating peaks and troughs,but the overall trend is decreasing.The year of 1992 is the most serious.In terms of the spatial distribution of the burned area,it gradually decreases from southeast to northwest,and is relatively dense in some areas(such as Longnan).According to the annual spatial distribution characteristics of the burned area,it can be seen that the frequency of repeated fires in the same area is low.(3)The contribution rates of different influencing factors are determined based on the Geodetector.The results show that the significant contribution factors are annual mean wind speed,annual mean air temperature,slope,relative humidity and altitude.(4)Through Sen+Mann-Kendall analysis,it is concluded that the forest vegetation in southern Jiangxi is mainly on the rise from 1990 to 2019,accounting for 95.10% of the whole area,and the area with a decreasing trend is only 4.90%.According to the trend of vegetation change before and after the fire,the vegetation decreased in 1.68%of the total area of the fire,and the intensity of vegetation change increased slightly before the fire.98.32% of the burned area shows an increasing trend.
Keywords/Search Tags:Burned area, Spatiotemporal variation, Google Earth Engine, Influencing factor, Vegetation trend
PDF Full Text Request
Related items