Photogrammetric reconstruction of declassified Corona imagery to establish a reference condition for a large river system: A case study of the Ganga River Basin, India.

The Ganga River basin sustains 43% of Indias population and faces significant damage due to anthropogenic stressors such as dams and barrages. To define and design a river management strategy that can sustain ecosystem services, a comprehensive understanding of the baseline geomorphic condition of the river is essential. Historical imageries like the declassified images of the Corona spy mission of the 1960s help to gain insights into the long-term temporal variability of the river channels where long-term monitoring datasets are unavailable and predates the major anthropogenic interventions along the Ganga. However, this dataset is difficult to use owing to significant image distortion and lack of information regarding the flight, film, and camera/sensor properties. Image to image georeferencing is time consuming and is still popularly used and the image distortion due to oblique angle of exposure and conical projection is often neglected. In this work, we used Structure-from-motion (SfM), an image-based photogrammetric method that can extract relative elevation data from digital images, overcoming the limitations of classical photogrammetry and georeferencing. We extracted point cloud and georeferenced it to create a Digital Surface Model (DSM) from each stereopair to generate orthophotos with positional errors < 15 metres which is small relative to the changes detected in the dynamic large river systems such as the river Ganga. We have further demonstrated the use of orthorectified Corona images to assess the channel planform changes in the last six decades by extracting the planform parameters using semi-automated methods and estimating the temporal changes in river channel width and planform morphology. We observed a two-to-three-fold decrease in channel width and a significant shift in channel position and channel planform changes relative to recent imageries. The study develops an understanding of the reference condition of the Ganga River using state-of-the-art techniques to document the alterations in river form, which can help river managers to design and implement strategies for river rehabilitation.