基于CALIOP 和MODIS 的北极地区海雾检测研究

极地地区的海雾给极地科考和海冰研究带来了挑战, 但由于相关监测数据较少, 因此对极地地区海雾的研究 还相对匮乏。基于 CALIOP 传感器可以观测垂直方向上云雾信息的特性, 使用准同步观测的 MODIS 中分辨率成像光 谱仪对北极地区的云雾信息进行分析。首先使用深度神经网络模型反演云顶高度, 再根据云高确定是否为海雾。并 且就不同波段对反演结果的影响进行了分析。结果显示使用深度神经网络反演的云顶高度平均绝对误差要比传统方 法的结果低 1774.280 m, 可以更好地对云顶高度进行反演, 提高了海雾检测精度。

Arctic sea fog detection using CALIOP and MODIS

Sea fog in polar regions poses a challenge to the research on polar science and sea ice. However, due to the lack of relevant cloud monitoring data in the polar region, the research on sea fog in polar regions is still relatively scare. Based on the CALIOP sensor′s ability to observe cloud information in the vertical direction, the MODIS medium resolution imaging spectrometer with plesiochronous observation is used to analyze cloud information in the Arctic region. Firstly, the deep neural network model is applied to invert the cloud top height. Then, according to the inverted cloud top height, whether it is sea fog can be ascertained. Furthermore, the influence of different wavebands on the inversion results is also analyzed. The results show that the average absolute error of the cloud top height inverted by the deep neural network is 1774.280 m lower than that of the traditional method, indicating that using deep neural network model can invert cloud top height better and more accurately, which can improve the detection accuracy of sea fog.