The impact of surface reflectance variability on total column differential absorption LiDAR measurements of atmospheric CO₂

Abstract. The remote sensing technique, total column differential absorption LiDAR (TC-DIAL) has been proposed in a number of feasibility studies as a suitable method for making total column measurements of atmospheric CO₂ from space. Among the sources of error associated with TC-DIAL retrievals from space is an undefined modulation of the received signals resulting from the variability in the Earth's surface reflectance between the LiDAR pulses. This source of uncertainty is investigated from a satellite perspective by the application of a computer model for spaceborne TC-DIAL instruments. The simulations are carried out over Europe and South America using modified MODIS surface reflectance maps and a DIAL configuration similar to that suggested for the proposed ESA A-SCOPE mission. A positive bias of 0.01 ppmv in both continental test sets is observed using 10 Hz pulse repetition frequency and 200 km integration distance. This bias is a consequence of non-linearity in the DIAL equation, and in particular regions such as the Alps and over certain coastlines it contributes to positive errors of between 0.05 and 0.16 ppmv for 200 and 50 km integration distances. These retrieval errors are defined as lower bound estimates owing to the likely resolution difference between the surface reflectance data and the expected surface heterogeneity observed by a DIAL instrument.