Developing and Testing a Deep Learning Approach for Mapping Retrogressive Thaw Slumps
Developing and Testing a Deep Learning Approach for Mapping Retrogressive Thaw Slumps
复制标题
DOI：
10.3390/rs13214294
复制
发表时间：
2021-10
期刊：
影响因子：
--
通讯作者：
Ingmar Nitze;Konrad Heidler;S. Barth;G. Grosse
中科院分区：
文献类型：
--
作者：
Ingmar Nitze;Konrad Heidler;S. Barth;G. Grosse
在变暖的北极，与永冻层有关的扰动，如退化融化滑塌（RTS），正变得越来越丰富和动态，对永冻层的稳定性和生物地球化学循环在当地到区域尺度上的严重影响。尽管最近在地球观测领域取得了进展，但其中许多仍然没有被发现，因为RTS是高度动态的，小的，分散在遥远的永久冻土区。在这里，我们使用PlanetScope卫星图像、ArcticDEM和辅助数据集评估了使用深度学习自动分割RTS的潜在优势和局限性。我们分析了在加拿大和俄罗斯的六个不同的受融化影响的地区进行泛北极升级和区域交叉验证的可转移性和潜力，并进行了独立的培训和验证。我们进一步测试了最先进的模型架构（UNet，UNet++，DeepLabv 3）和编码器网络，以找到最佳的模型配置，以便扩展到大陆规模。最好的深度学习模型在六个地区中的四个地区取得了从良好到非常好的结果（maxIoU：0.39到0.58; Lena River，Horton Delta，Herschel Island，Kolguev Island），而在两个地区（Banks Island，Tuktoyaktuk）失败。在测试的架构中，UNet++表现最好。区域性能的巨大差异突出了在不同的环境条件下，用于分割不同永久冻土景观RTS的训练数据中需要足够的数量，质量和空间变异性。通过我们高度自动化和可配置的工作流程，我们看到了转移到活动RTS集群的巨大潜力（例如，Peel Plateau）和升级到更大的区域。
In a warming Arctic, permafrost-related disturbances, such as retrogressive thaw slumps (RTS), are becoming more abundant and dynamic, with serious implications for permafrost stability and bio-geochemical cycles on local to regional scales. Despite recent advances in the field of earth observation, many of these have remained undetected as RTS are highly dynamic, small, and scattered across the remote permafrost region. Here, we assessed the potential strengths and limitations of using deep learning for the automatic segmentation of RTS using PlanetScope satellite imagery, ArcticDEM and auxiliary datasets. We analyzed the transferability and potential for pan-Arctic upscaling and regional cross-validation, with independent training and validation regions, in six different thaw slump-affected regions in Canada and Russia. We further tested state-of-the-art model architectures (UNet, UNet++, DeepLabv3) and encoder networks to find optimal model configurations for potential upscaling to continental scales. The best deep learning models achieved mixed results from good to very good agreement in four of the six regions (maxIoU: 0.39 to 0.58; Lena River, Horton Delta, Herschel Island, Kolguev Island), while they failed in two regions (Banks Island, Tuktoyaktuk). Of the tested architectures, UNet++ performed the best. The large variance in regional performance highlights the requirement for a sufficient quantity, quality and spatial variability in the training data used for segmenting RTS across diverse permafrost landscapes, in varying environmental conditions. With our highly automated and configurable workflow, we see great potential for the transfer to active RTS clusters (e.g., Peel Plateau) and upscaling to much larger regions.