METHANE EMISSIONS ON LARGE SCALES

Two separate studies have been undertaken to improve estimates of meth ane emissions on a landscape scale. The first study took place over a palsa mire in northern Finland in August 1995. A tethered balloon and a tunable diode laser were used to measure profiles of methane in the nocturnal boundary layer. Using a simple box method or the flux gradie nt technique fluxes ranging from 18.5 to 658 mu mol m(-2) h(-1) were c alculated. The large fluxes may be caused by advection of methane pock ets across the measurement site, reflecting the heterogeneous nature o f methane source strengths in the surrounding area. Under suitable con ditions, comparison with nearby ground-based eddy-correlation results suggested that the balloon techniques could successfully measure fluxe s on field scales. The second study was carried out by the NERC Scient ific Services Atmospheric Research Airborne Support Facility using the Hercules C130 operated by the United Kingdom Meteorological Research Flight. A flight path around the northern coastline of Britain under s teady West-East wind conditions enabled the measurement of methane con centrations up- and down-wind of northern Britain. Using a simple one- dimensional, constant-source diffusion model, the difference between t he upwind and downwind concentrations was accounted for by methane emi ssion from the surface. The contribution to methane emissions from liv estock was also modelled. Modelled non-agricultural methane emissions decreased with increasing latitude with fluxes in northern England bei ng a factor of 4 greater than those in northern Scotland. Since the on ly major methane source in northern Scotland was peat bogs, these resu lts indicated that emissions over northern England were dominated by a nthropogenic sources. Emissions from livestock accounted for 12% of th e total flux over northern England, decreasing to 4% in southern Scotl and and becoming negligible in northern Scotland. The total methane fl ux over northern Scotland was consistent with previous results from th e area, indicating that this method of data analysis provided good est imates of large scale methane emissions. (C) 1998 Elsevier Science Ltd . All rights reserved.