Automated Geolocation Design for Nanosatellite Missions Missions Based on Coastline Detection for Near Infrared Spectrometers

Geolocation, defined as the determination of the location on the ground from which a remote sensing observation is made, is a critical task in delivering accurate and reliable scientific data from an Earth observation (EO) spaceborne measurement system. Geolocation accuracy is a mission requirement, whose validation in EO missions with low resolution spectrometers is a challenging problem. While various techniques are proposed for large space missions, limited techniques are available for CubeSat-class, resource constrained, low resolution spectrometer payloads. In this dissertation, a novel approach for automated geolocation accuracy assessment using coastline detection was developed and implemented for CubeSat-class nanosatellite missions with low-resolution single-pixel spectrometers in nadir viewing configurations. The algorithm was demonstrated using Argus 1000 near-infrared (NIR) spectrometer data aboard the CanX-2 nanosatellite in order to verify the mission requirements of 1 (1) pointing knowledge uncertainty. The same algorithm was also applied to airborne data using the Ocean Optics FLAME-NIR spectrometer for validation purposes. Radiometric calibration was also performed on both instruments for use in field campaigns.

From the spaceborne data analysis, 270 spectrometer data sets were analyzed, from which 55 coastlines were detected. The mean angular error in the data sets was 0.44 and the standard deviation was 0.57, which was consistent with the CanX-2 mission requirements. The airborne data analysis yielded similar results using the same coastline detection algorithm as the spaceborne data analysis. A total of 7 data sets were acquired over two days at two data collection sites. 36 coastline crossings were detected and a mean angular error of 9.7 was observed, with a standard deviation of 6.8. Compared to the spaceborne data, higher error was observed due to the lower speed and altitude of the UAV compared to the satellite.