| Remote sensing satellite missions involve the use of Earth observing satellite payloads, whose pointing location on the surface of Earth must be accurately determined using a process called geolocation. This thesis details the implementation of a geolocation algorithm for spectral data from the Argus 1000 spectrometer, a nadir viewing instrument aboard the CanX-2 nanosatellite. Additionally, this thesis examines the accuracy of the geolocation algorithm and examines several methods for assessing and reducing different sources of geolocation uncertainty. These methods include an extended Kalman filter for attitude knowledge, optical ranging for satellite position, and coastline detection for on-board timing and attitude control. |
