Abstract:
This work deals with radar satellite imagery and its potential to assist of humanitarian operations. As the number of displaced people annually increases, both hosting countries and relief organizations face new challenges which are often related to unclear situations and lack of information on the number and location of people in need, as well as their environments. It was demonstrated in numerous studies that methods of earth observation can deliver this important information for the management of crises, the organization of refugee camps, and the mapping of environmental resources and natural hazards. However, most of these studies make use of -high-resolution optical imagery, while the role of radar satellites is widely neglected. At the same time, radar sensors have characteristics which make them highly suitable for humanitarian response, their potential to capture images through cloud cover and at night in the first place. Consequently, they potentially allow quicker response in cases of emergencies than optical imagery.
This work demonstrates the currently unused potential of radar imagery for the assistance of humanitarian operations by case studies which cover the information needs of specific emergency situations. They are thematically grouped into topics related to population, natural hazards and the environment. Furthermore, the case studies address different levels of scientific objectives: The main intention is the development of innovative techniques of digital image processing and geospatial analysis as an answer on the identified existing research gaps. For this reason, novel approaches are presented on the mapping of refugee camps and urban areas, the allocation of biomass and environmental impact assessment. Secondly, existing methods developed for radar imagery are applied, refined, or adapted to specifically demonstrate their benefit in a humanitarian context. This is done for the monitoring of camp growth, the assessment of damages in cities affected by civil war, and the derivation of areas vulnerable to flooding or sea-surface changes. Lastly, to foster the integration of radar images into existing operational workflows of humanitarian data analysis, technically simple and easily-adaptable approaches are suggested for the mapping of rural areas for vaccination campaigns, the identification of changes within and around refugee camps, and the assessment of suitable locations for groundwater drillings.
While the studies provide different levels of technical complexity and novelty, they all show that radar imagery can largely contribute to the provision of a variety of information which is required to make solid decisions and to effectively provide help in humanitarian operations. This work furthermore demonstrates that radar images are more than just an alternative image source for areas heavily affected by cloud cover. In fact, what makes them valuable is their information content regarding the characteristics of surfaces, such as shape, orientation, roughness, size, height, moisture, or conductivity. All these give decisive insights about man-made and natural environments in emergency situations and cannot be provided by optical images
Finally, the findings of the case studies are put into a larger context, discussing the observed potential and limitations of the presented approaches. The major challenges are summarized which need be addressed to make radar imagery more useful in humanitarian operations in the context of upcoming technical developments. New radar satellites and technological progress in the fields of machine learning and cloud computing will bring new opportunities. At the same time, this work demonstrated the large need for further research, as well as for the collaboration and transfer of knowledge and experiences between scientists, users and relief workers in the field. It is the first extensive scientific compilation of this topic and the first step for a sustainable integration of radar imagery into operational frameworks to assist humanitarian work and to contribute to a more efficient provision of help to those in need.