Earth Polychromatic Imaging Camera Geolocation; Strategies to Reduce Uncertainty

Karin Blank,Jay Herman,Alexander Marshak,Liang-Kang Huang

doi:10.3389/frsen.2021.715296

Abstract

Earth Polychromatic Imaging Camera occupies a unique point of view for an Earth imager by being located approximately 1.5 million km from the planet at Earth-Sun Lagrange point, L1. This creates a number of unique challenges in geolocation, some of which are distance and mission specific. To solve these problems, algorithmic adaptations need to be made for calculations used for standard geolocation solutions, as well as artificial intelligence-based corrections for star tracker attitude and optical issues. This paper discusses methods for resolving these issues and bringing the geolocation solution to within requirements.

Highlights

InstrumentThe Earth Polychromatic Imaging Camera (EPIC) is an instrument on the Deep Space Climate Observatory (DSCOVR), which orbits the earth-sun lagrange 1 (L1) Earth-Sun Lagrange point
The likely result of this is a lift of several millimeters of the charge-coupled device (CCD) at the lower right corner
It is difficult to ascertain when exactly this happened but based on reviews of work orders performed on the reassembly of the instrument, as well as due to some other issues witnessed with the CCD not seen during ground testing, the probability is that it happened during launch

Summary

Introduction

The Earth Polychromatic Imaging Camera (EPIC) is an instrument on the Deep Space Climate Observatory (DSCOVR), which orbits the L1 Earth-Sun Lagrange point. As an Earth viewing instrument, it has a unique view of the planet, taking 13–21 images daily at local noon. The instrument is a 30 cm Cassegrain telescope with 2048x2048 charge-coupled device (CCD), using two filter wheels, and containing a set of 10 bands at wavelengths between 317 and 780 nm (Figure 1) (DSCOVR:EPIC, 2016). A typical imaging session consists of 10 images taken once of each band. In creating the science products, the images are calibrated into units of counts/second and geolocation is calculated (Marshak et al, 2018).

Methods

Results

Conclusion