Environmental Earth Sciences - Call for Papers

Deep learning for earth resource and environmental remote sensing

Guest Editors: Carlos Enrique Montenegro Marin, Xuyun Zhang, Nallappan Gunasekaran

In the current era, climate change is the biggest problem mankind is facing, which forms several other environmental consequences. Earth resource and environmental monitoring are fundamental for the preservation of a proper environment in the long term. It is an efficient tool to assess environmental conditions and natural resources supporting sustainable policy developments, regulatory measures, and their implementation enriching the ecosystem positively. Remote sensing provides essential data that assists in monitoring various aspects of the environment such as land cover classification, change detection, wildlife management, forest fire detection, discovery and mapping of rugged topography of the ocean floors, etc. Overall, it presents a global perspective and a plethora of data about the earth systems. Remotely sensed data could be of spectral, temporal, and spatial resolutions.  Remote sensing data obtained from earth resources and satellites require processing before it is efficiently used by the researchers and the earth scientists. The finer the resolution, the more details can be extracted from the images. The major concern here is that it is often difficult to combine all of the desirable features (spatial, temporal, spectral) in a single remote sensor. In short, interpreting satellite and sensor data is not an easy task. Hence, researchers need to make trade-offs, and they should carefully understand and select the type of data that is suitable for any given scenario of study. This is where exactly the role of deep learning comes into the picture for remote sensing applications. 

Over many decades, researchers have been actively involved in applying various algorithmic approaches to resolve the challenges of interpreting earth images. The intersection of deep learning and environmental remote sensing creates massive opportunities that were not possible before. Broadly speaking, deep learning can create a revolution in various earth resource and environmental remote sensing applications opening up an entirely novel frontier of tools and algorithms useful for earth monitoring with state-of-the-art results. This is because the potential of deep learning for earth imagery is immense and would grow rapidly so that more avenues can be explored. This includes air monitoring, water monitoring, waste monitoring, biodiversity conservation, etc. It provides complementary information to effectively meet the requirements of ground-level environmental remote sensing applications with unprecedented levels of accuracy. However, with the growing revolution in big data, applying deep learning techniques for earth observation can be a little challenging for even basic workflows. Thus, to effectively address these concerns and leverage deep learning advances, this thematic issue invites cutting-edge research works on deep learning for earth resource and environmental remote sensing applications.

The topic of interest includes the following:

• Deep learning for land cover and land change analysis
• Innovations in deep learning for UAV remote sensing and real-time processing
• Deep learning assisted coastal zone remote sensing for environmental sustainability
• Data processing, interpretation, and analysis of environmental resources with deep learning algorithms
• Deep learning for vegetation status monitoring and applications
• Challenges in intersecting deep learning with earth resources management and its appropriate solutions
• Deep learning and computer vision applications to geospatial analysis
• Role of deep learning in disaster mitigation planning and recovery
• Deep learning assisted remote sensing for wildlife management and biodiversity conservation
• Moving forward with satellite imagery and deep learning applications from a future perspective