VADUGS: a neural network for the remote sensing of volcanic ash with MSG/SEVIRI trained with synthetic thermal satellite observations simulated with a radiative transfer model
VADUGS: a neural network for the remote sensing of volcanic ash with MSG/SEVIRI trained with synthetic thermal satellite observations simulated with a radiative transfer model
复制标题
DOI：
10.5194/nhess-22-1029-2022
复制
发表时间：
2022-03
影响因子：
4.6
通讯作者：
L. Bugliaro;Dennis Piontek;S. Kox;Marius Schmidl;B. Mayer;Richard Müller;M. Vazquez-Navarro;D. Peters;R. Grainger;J. Gasteiger;J. Kar
中科院分区：
文献类型：
--
作者：
L. Bugliaro;Dennis Piontek;S. Kox;Marius Schmidl;B. Mayer;Richard Müller;M. Vazquez-Navarro;D. Peters;R. Grainger;J. Gasteiger;J. Kar
抽象的。在世界各地火山喷发后，由于火山灰对空中交通构成威胁，因此监测火山灰在大气中的扩散情况是卫星遥感的一项重要任务。在这项工作中，我们提出了一种新的方法，该方法是为Eyjafjallajökull火山灰量身定做的，但也适用于其他喷发，它使用地球静止气象卫星第二代卫星上的SEVIRI成像仪的热观测来探测火山灰云，并在白天和夜间确定它们的质量、柱浓度和顶部高度。这一方法需要汇编大量的SEVIRI合成观测数据集，以训练人工神经网络。这是通过RTSIM工具完成的，该工具结合了大气、表面和火山灰的属性，并自动运行整个SEVIRI圆盘的大量辐射传递计算。由此产生的算法被称为“VADUGS”(使用地球同步卫星探测火山灰)，并已通过独立的辐射传输模拟进行了评估。VADUGS检测被灰尘污染的像素，检测概率为0.84，误警率为0.05。VADUGS提供的灰柱浓度相关系数高达0.5，对于浓度小于2.0g m−2的浓度，散布高达0.6g m−2，对于中等视角35-65∘，高估在5%-50%之间，但对于接近90或0∘的卫星观测天顶角，高估高达300%。火山灰顶部高度主要被低估，对于40到50−之间的观测天顶角，最小的低估∘为9%。对于质量柱浓度较高的灰云，绝对误差小于70%，相关系数高达0.7。与CALIPSO/CALIOP的空间激光雷达观测结果的比较证实了这些结果：对于2011年6月从普耶霍-科隆·考勒火山产生的火山灰云上方的六个立交桥，VADUGS显示出与相应的激光雷达数据相似的特征，相关系数为0.49，对火山灰柱浓度的高估为55%，尽管仍在CALIOP的不确定范围内。与另一种火山灰算法的比较表明，这两种方法都提供了可信的探测结果，VADUGS能够探测到距离艾亚菲亚德拉火山更远的火山灰，但有时会错过靠近喷口的厚厚的火山灰云。VADUGS在德国气象局运行，该应用程序也被介绍。
Abstract. After the eruption of volcanoes around the world, monitoring of the dispersion of ash in the atmosphere is an important task for satellite remote sensing since ash represents a threat to air traffic. In this work we present a novel method, tailored for Eyjafjallajökull ash but applicable to other eruptions as well, that uses thermal observations of the SEVIRI imager aboard the geostationary Meteosat Second Generation satellite to detect ash clouds and determine their mass column concentration and top height during the day and night. This approach requires the compilation of an extensive data set of synthetic SEVIRI observations to train an artificial neural network. This is done by means of the RTSIM tool that combines atmospheric, surface and ash properties and runs automatically a large number of radiative transfer calculations for the entire SEVIRI disk. The resulting algorithm is called “VADUGS” (Volcanic Ash Detection Using Geostationary Satellites) and has been evaluated against independent radiative transfer simulations. VADUGS detects ash-contaminated pixels with a probability of detection of 0.84 and a false-alarm rate of 0.05. Ash column concentrations are provided by VADUGS with correlations up to 0.5, a scatter up to 0.6 g m−2 for concentrations smaller than 2.0 g m−2 and small overestimations in the range 5 %–50 % for moderate viewing angles 35–65∘, but up to 300 % for satellite viewing zenith angles close to 90 or 0∘. Ash top heights are mainly underestimated, with the smallest underestimation of −9 % for viewing zenith angles between 40 and 50∘. Absolute errors are smaller than 70 % and with high correlation coefficients of up to 0.7 for ash clouds with high mass column concentrations. A comparison with spaceborne lidar observations by CALIPSO/CALIOP confirms these results: For six overpasses over the ash cloud from the Puyehue-Cordón Caulle volcano in June 2011, VADUGS shows similar features as the corresponding lidar data, with a correlation coefficient of 0.49 and an overestimation of ash column concentration by 55 %, although still in the range of uncertainty of CALIOP. A comparison with another ash algorithm shows that both retrievals provide plausible detection results, with VADUGS being able to detect ash further away from the Eyjafjallajökull volcano, but sometimes missing the thick ash clouds close to the vent. VADUGS is run operationally at the German Weather Service and this application is also presented.