We present a general methodology and first results of total column trace gas retrievals from hyperspectral infrared observations. More specifically, we made use of a physical scheme to retrieve methane (CH4) profiles from atmospheric infrared sounder (AIRS) measurements. The method firstly required a forward calculation that maps the a priori state vector into measurement space, and for this we used the fast forward model, SARTA. Secondly, the calculation was inverted to derive at a posterior state of the atmosphere. We used gas profiles from the atmospheric chemistry experiment (ACE) together with ECMWF model fields for the assimilation of the state vector. ACE profiles were also used in the validation procedure. The retrievals were improved by up to 5% at the surface. Following the same scheme, we retrieved total column concentration of CH4 for an entire AIRS granule, thereby demonstrating its potential for use in operational air quality monitoring schemes at regional scales.