Datasets for Deep Learning for Classifying and Characterizing Atmospheric Ducting Within the Maritime Setting

Propagation factors are sparsely sampled along a horizontal flight path, at an altitude of 6m and range between 20 and 50km, within the rectangular 2D problem domain described in the associated article. We consider evaporation ducting environments with varying duct heights as well as surface-based ducting environments with varying combinations of base height, duct thickness, and M-deficit. Datasets are generated using the open source software, PETOOL, by Ozgun et al. 2011 (doi: 10.17632/yh42r43cpy.1). Code for processing data and training/evaluating the deep learning model can be found at: https://github.com/nonlinearfun/deep-learning-em-ducting本数据集在相关论文所述的矩形二维问题域内，于6米高度、20至50千米的距离范围内，沿水平飞行路径对传播因子（propagation factors）进行稀疏采样。本研究涵盖波导高度各异的蒸发波导（evaporation ducting）环境，以及基准高度、波导厚度与M亏损（M-deficit）组合各不相同的地面基波导（surface-based ducting）环境。本数据集由Ozgun等人2011年开发的开源软件PETOOL生成（DOI：10.17632/yh42r43cpy.1）。用于数据处理及深度学习模型训练与评估的代码可在以下网址获取：https://github.com/nonlinearfun/deep-learning-em-ducting