基于时空融合技术的森林火灾遥感动态监测  

Remote sensing dynamic monitoring of forest fire based on spatiotemporal fusion technology

作　　者：黄武彪 栾海军 李大成[4] HUANG Wubiao;LUAN Haijun;LI Dacheng(College of Computer and Information Engineering,Xiamen University of Technology,Xiamen 361024,China;College of Geological Engineering and Geomatics,Chang'an University,Xi'an 710064,China;Big Data Institute of Digital Natural Disaster Monitoring in Fujian,Xiamen University of Technology,Xiamen 361024,China;College of Mining Engineering,Taiyuan University of Technology,Taiyuan 030024,China)

机构地区：[1]厦门理工学院计算机与信息工程学院,福建厦门361024 [2]长安大学地质工程与测绘学院,陕西西安710064 [3]厦门理工学院数字福建自然灾害监测大数据研究所,福建厦门361024 [4]太原理工大学矿业工程学院,山西太原030024

出　　处：《自然灾害学报》2022年第1期265-276,共12页Journal of Natural Disasters

基　　金：国家自然科学基金项目(41601350);福建省自然科学基金项目(2020J01261);厦门理工学院科学技术研究项目(“科研攀登计划”资助项目)(XPDKT19010)。

摘　　要：针对单一时空融合方法或使用单一中等空间分辨率影像(如Landsat影像)和MODIS影像时空融合的不足,本文提出综合利用经典的STARFM算法、基于地物内组分时相变化模型的地表反射率时空融合算法,联合使用多种空间分辨率更优(≤30 m)的传感器影像,以“时间最邻近及空间分辨率优先”为原则对传统单一中等空间分辨率影像预测周期(如Landsat影像为16天)进行分段独立预测,并优化组合两种预测方法的预测结果,进而获取更为精确的逐日中等空间分辨率预测影像。基于上述方法所得结果,可应用于森林火灾监测场景中。以四川凉山木里县3·30森林大火为例,综合利用MOD09GA、Landsat8 OLI、Sentinel-2、GF-1 WFV遥感影像数据进行实验研究,基于预测所得逐日中等空间分辨率影像提取火灾指标因子(燃烧面积指数和归一化燃烧指数),分析森林火灾演化态势。结果表明:(1)多类型中高空间分辨率遥感影像的综合利用,有利于解决传统单一中等空间分辨率影像预测周期跨度过长、总体精确度低的问题,可获取更为精确的逐日中等空间分辨率预测影像;(2)两种算法在不同类型遥感数据融合应用中各有其局限性,两种方法联立使用具有理论价值与实际意义;(3)基于时空融合影像分析火灾演化态势时,归一化燃烧指数计算结果更敏感、更有效。研究认为,基于时空融合技术的森林火灾遥感动态监测具有可行性,具有进一步深入研究的价值与意义。The single spatiotemporal fusion method or the use of a single medium spatial resolution image(such as Landsat image)and MODIS image spatiotemporal fusion have shortcomings,this paper proposes a comprehensive use of the classic Spatial and Temporal Adaptive Reflectance Fusion Model algorithm and the surface reflectance space-time fusion algorithm based on the temporal phase change model of the components within the ground feature,combined with various sensor images with better spatial resolution(≤30 m),based on the principle of "time closest and spatial resolution priority" to traditional single medium spatial resolution image prediction period(such as Landsat image is 16 days)to segmented independent prediction,and optimize the combination of the prediction results of the two methods to obtain more accurate day-to-day medium spatial resolution prediction images. Based on the results obtained from the above method,it can be applied to forest fire monitoring scenarios. Taking the 3·30 forest fire in Muli County,Liangshan City,Sichuan Province as an example,comprehensively utilize MOD09 GA,Landsat8 OLI,Sentinel-2,GF-1 WFV remote sensing image data for experimental research,and extract fire index factors(Burning area index and normalized burning index)based on the daily medium spatial resolution images—obtained from predictions,and analyzed the evolution of forest fires. The results show that:a)The comprehensive utilization of various types of medium and high spatial resolution remote sensing images can solve the problems of long prediction period and low accuracy of traditional single medium spatial resolution images,and can obtain more accurate daily medium spatial resolution prediction images;b)The two algorithms have their own limitations in different types of remote sensing data fusion applications,the combined use of the two methods has theoretical value and practical significance;c)The results of Normalized Burn Ratio are more sensitive and effective when analyzing fire evolution trend based on spatiotem

关 键 词：森林火灾 遥感动态监测 时空融合 时空自适应反射率融合模型 地物组分 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] X43[自动化与计算机技术—控制科学与工程]