Quantifying burned area of wildfires in the western United States from polar-orbiting and geostationary satellite active-fire detections
Quantifying burned area of wildfires in the western United States from polar-orbiting and geostationary satellite active-fire detections
复制标题
通过极轨和地球静止卫星主动火灾探测来量化美国西部野火的烧毁面积
DOI：
10.1071/wf22022
复制
发表时间：
2023
影响因子：
3.1
通讯作者：
Saide, Pablo E.
中科院分区：
文献类型：
--
作者：
Berman, Melinda T.;Ye, Xinxin;Thapa, Laura H.;Peterson, David A.;Hyer, Edward J.;Soja, Amber J.;Gargulinski, Emily M.;Csiszar, Ivan;Schmidt, Christopher C.;Saide, Pablo E.
卫星准确估算燃烧面积是改进生物质燃烧排放模型、研究火灾演变和评估环境影响的关键。以前的研究发现，目前从卫星活动火灾数据估计火灾燃烧面积的方法并不总是提供准确的估计。在这项工作中，我们开发了一种新的算法，该算法基于包含累积可见光红外成像辐射计套件（VIIRS）主动火灾探测的非凸多边形边界的面积来估计每小时累积烧伤面积。每小时的时间序列是通过将VIIRS估计值与来自GOES-17（地球静止运行环境卫星）数据的火力辐射功率（FRP）估计值相结合而产生的。结论、关键结果和意义我们评估了该算法在航空观测中燃烧面积累积和变化的性能，并特别检查了控制边界向区域多边形内部收缩多少的参数选择的敏感性。2019 - 2020年多起火灾的逐时累积燃烧面积与美国林务局国家红外业务（NIROPS）收集的机载红外（IR）观测结果普遍具有很强的相关性，相关系数值通常大于0.95，误差<20%。
Background Accurately estimating burned area from satellites is key to improving biomass burning emission models, studying fire evolution and assessing environmental impacts. Previous studies have found that current methods for estimating burned area of fires from satellite active-fire data do not always provide an accurate estimate.Aims and methods In this work, we develop a novel algorithm to estimate hourly accumulated burned area based on the area from boundaries of non-convex polygons containing the accumulated Visible Infrared Imaging Radiometer Suite (VIIRS) active-fire detections. Hourly time series are created by combining VIIRS estimates with Fire Radiative Power (FRP) estimates from GOES-17 (Geostationary Operational Environmental Satellite) data.Conclusions, key results and implication We evaluate the performance of the algorithm for both accumulated and change in burned area between airborne observations, and specifically examine sensitivity to the choice of the parameter controlling how much the boundary can shrink towards the interior of the area polygon. Results of the hourly accumulation of burned area for multiple fires from 2019 to 2020 generally correlate strongly with airborne infrared (IR) observations collected by the United States Forest Service National Infrared Operations (NIROPS), exhibiting correlation coefficient values usually greater than 0.95 and errors <20%.