Extracting Bolide Light Curves from GOES GLM Data

NASA's Asteroid Threat Assessment Project (ATAP) aims to produce world-leading quantitative estimates of the asteroid impact threat. High-quality models of bolide fragmentation and energy deposition are key to this effort and fitting existing models to observations can provide validation and illuminate shortcomings. Data from the Geostationary Lightning Mapper (GLM) instruments on GOES 16 and 17 satellites have been shown to contain the signatures of many independently confirmed bolides, often from both viewing angles simultaneously, and constitute an important untapped source of observational data. The design of the GLM instruments and their onboard data processing algorithms, which are tailored to monitor lightning, pose a number of significant challenges to recovering calibrated bolide light curves. In particular the adaptive thresholding and background estimation algorithms can cause a significant fraction of the measured bolide flux to be discarded. We present a method for extracting high-quality bolide light curves with associated uncertainties from GLM data. At the heart of our method is a model of image formation which we fit to the available data for a known bolide. The algorithm effectively fills in missing data by leveraging what we know about the instrument's optics and the onboard data processing. We demonstrate its effectiveness by applying it to selected cases where independent observations are available.