The CO2Image mission: retrieval studies and performance analysis

Abstract

The CO2Image satellite mission, led by the German Aerospace Center (DLR), aims to demonstrate the feasibility of quantifying carbon dioxide (CO2) and methane (CH4) emissions from medium-size point sources. Several DLR institutes are currently working on the reliminary design phase (Phase B) of the mission. Here we present a performance analysis based on the current instrument specifications. The Beer InfraRed Retrieval Algorithm (BIRRA), the line-by-line radiative transfer model Py4CAtS (Python for Computational ATmospheric Spectroscopy) and a COSIS (Carbon dioxide Sensing Imaging Spectrometer) instrument model are employed to infer CO2 and CH4 concentrations from synthetic COSIS spectra. We evaluate the instrument's performance and determine if it meets the intended requirements. The study assesses uncertainties in the retrieved concentrations as well as errors in point source emission estimates caused by instrument noise. First results suggest that the detection and quantification limits stated in the mission requirements document are justified. The analysis also demonstrates that retrieval errors tend to increase when the signal-to-noise ratio is low, complicating the distinction between emission sources and background concentrations. Furthermore, we discuss non-instrumental effects and demonstrate that the fit quality significantly improves if a low-level plume is scaled instead of a background reference profile that covers the atmosphere's full vertical extent. The analysis on heterogeneous scenes (high albedo contrast) reveals that the various instrument setups perform similarly for both molecules.