Abstract
Earth Polychromatic Imaging Camera (EPIC) raw level-0 (L0) data in one channel is a 12-bit 2,048 × 2,048 pixels image array plus auxiliary data such as telemetry, temperature, etc. The EPIC L1a processor applies a series of correction steps on the L0 data to convert them into corrected count rates (level-1a or L1a data): Dark correction, Enhanced pixel detection, Read wave correction, Latency correction, Non-linearity correction, Temperature correction, Conversion to count rates, Flat fielding, and Stray light correction. L1a images should have all instrumental effects removed and only need to be multiplied by one single number for each wavelength to convert counts to radiances, which are the basis for all higher-level EPIC products, such as ozone and sulfur dioxide total column amounts, vegetation index, cloud, aerosol, ocean surface, and vegetation properties, etc. This paper gives an overview of the mathematics and the pre-launch and on-orbit calibration behind each correction step.
Highlights
The Earth Polychromatic Imaging Camera (EPIC) operates aboard the Deep Space Climate Observatory (DSCOVR) satellite that is orbiting the Sun at the Lagrange-1 point, L1, about 1.5 million kilometers away from Earth (Marshak et al, 2018)
Since filter wheel 1 is farther away from the detector, the diameter of the ghost image is larger for filters 1 to 5 than for filters 6 to 10 (Figure 10). Based on these measurements we developed a point spread function (PSF)-model for EPIC
The produced L1a data are corrected for all known instrumental effects and only need to be multiplied by a single number for each filter to obtain absolute calibrated radiances from the count rate
Summary
The Earth Polychromatic Imaging Camera (EPIC) operates aboard the Deep Space Climate Observatory (DSCOVR) satellite that is orbiting the Sun at the Lagrange-1 point, L1, about 1.5 million kilometers away from Earth (Marshak et al, 2018). Since it would go beyond the limits of this paper to describe each correction step in full detail, we focus on those corrections with more significant effects on the data, i.e., dark-count, flat-field, and stray light corrections.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call