The task of remote sensing is to measure the concentrations of specific trace gases in the atmosphere, and particularly in the troposphere. The presence of a species in the atmosphere can be detected through the presence of characteristic absorption features in the spectrum of light passing through the atmosphere, and the concentration of the species from the amount of absorption that takes place (in accordance with the Beer-Lambert Law).

The process of data retrieval described in this section takes the earthshine spectrum measured at the satellite for each ground pixel scanned and derives a value for the concentration of NO₂ in the column of troposphere above that pixel.

This section describes the procedures used to derive the concentrations of tropospheric NO₂ from earthshine spectra.

• Irradiance and earthshine
  The nature of the solar irradiance — the incident solar radiation before it encounters the Earth's atmosphere — and the use of its spectrum as a reference against which we can analyse the earthshine spectrum.
• Earthshine interactions
  The absorption signature of NO₂ in the earthshine spectrum has to be extracted from the many spectral features that are not the result of NO₂ absorption.
• DOAS
  The concentrations of NO₂ and other trace gases are extracted using a technique known as Differential Optical Absorption Spectroscopy, or DOAS. This section presents a graphical and a mathematical treatment of the DOAS technique.
• SCD_total to VCD_trop
  Our focus of interest is the relatively short 'vertical column' of troposphere directly above the ground pixel. The DOAS technique yields NO₂ concentrations for the 'total slant column', a long, slanting light path that depends on the solar zenith angle and the line of sight of the spectrometer towards the ground pixel. We need to process the total slant column density to arrive at values for the tropospheric vertical column density of NO₂.
• Summary
  This section, intended as a revision element, is a step-by-step presentation of all the major retrieval processes.