Abstract
<jats:p>Abstract. Methane (CH4) is the second most important greenhouse gas with an atmospheric lifetime of approximately 10 years. Due to its relatively short lifetime, it offers substantial mitigation potential with reduction in emissions. Lakes and wetlands constitute major natural CH4 sources, and offshore oil and gas production is a dominant anthropogenic source over open water. However, passive remote sensors have observational gaps over water due to low surface reflectance and challenging sun-glint measurement geometry, and in-situ observations also remain sparse. Active remote sensing using CO2 and CH4 Atmospheric Remote Monitoring – Flugzeug (CHARM-F) Integrated Path Differential Absorption (IPDA) lidar provides column-averaged methane mixing ratio (XCH4) over water bodies and wetlands independent of solar illumination and largely independent of surface reflectance. However, understanding how the lidar signal to noise ratio (SNR) varies with different surface types, and how this variability propagates into retrieval uncertainties is critical to fully exploit the potential of IPDA lidar over water bodies and wetlands. In this study, we analyze airborne CHARM-F IPDA lidar observations acquired during two Carbon dioxide and Methane (CoMet) field campaigns over heterogeneous land and water surfaces in Europe and North America. This lidar serves as the airborne demonstrator for the future space borne MEthane Remote sensing Lidar missioN (MERLIN) which will deliver global methane emission maps and reduce uncertainties in methane emission estimates. Here, we study the averaging biases of the lidar XCH4 retrievals and propagate SNR-derived uncertainties to the final XCH4 estimates building on the work of Tellier et al. (2018). An independent analysis of surface reflectance measurements from the hyperspectral imaging spectrometer of the Munich Aerosol Cloud Scanner (specMACS) against CHARM-F signal strength and SNR provides a direct observational link between surface properties, signal statistics, and retrieval performance. The lidar signals over dry land exhibit Gaussian distributions associated with high surface reflectance, high SNR and low non-correctable residual XCH4 uncertainties below 1 ppb. In contrast, wetlands and open water surfaces display weaker and broader signal distributions with SNR reduced by a factor of about 2 as compared to dry land. However, the retrieval uncertainty over water bodies remains below 1 ppb when using an appropriate water mask and SNR-weighted averaging of the lidar signals. These findings provide the first observationally constrained characterization of how surface type is related to lidar SNR, averaging bias, and retrieval uncertainty in airborne IPDA methane measurements. The results highlight the capability of IPDA to robustly measure methane over water, relevant for the MERLIN mission.</jats:p>