A31B-0045
Mapping pan-Arctic CH4 emissions using an adjoint method by integrating process-based wetland and lake biogeochemical models and atmospheric CH4 concentrations
Wednesday, 16 December 2015
Poster Hall (Moscone South)
Zeli Tan1, Qianlai Zhuang1, Daven K Henze2, Christian Frankenberg3, Edward J Dlugokencky4, Colm Sweeney5 and Alexander J. Turner6, (1)Purdue University, West Lafayette, IN, United States, (2)University of Colorado at Boulder, Boulder, CO, United States, (3)NASA Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, United States, (4)NOAA Boulder, Boulder, CO, United States, (5)NOAA Boulder, ESRL, Boulder, CO, United States, (6)Harvard University, Cambridge, MA, United States
Abstract:
Understanding CH4 emissions from wetlands and lakes are critical for the estimation of Arctic carbon balance under fast warming climatic conditions. To date, our knowledge about these two CH4 sources is almost solely built on the upscaling of discontinuous measurements in limited areas to the whole region. Many studies indicated that, the controls of CH4 emissions from wetlands and lakes including soil moisture, lake morphology and substrate content and quality are notoriously heterogeneous, thus the accuracy of those simple estimates could be questionable. Here we apply a high spatial resolution atmospheric inverse model (nested-grid GEOS-Chem Adjoint) over the Arctic by integrating SCIAMACHY and NOAA/ESRL CH4 measurements to constrain the CH4 emissions estimated with process-based wetland and lake biogeochemical models. Our modeling experiments using different wetland CH4 emission schemes and satellite and surface measurements show that the total amount of CH4 emitted from the Arctic wetlands is well constrained, but the spatial distribution of CH4 emissions is sensitive to priors. For CH4 emissions from lakes, our high-resolution inversion shows that the models overestimate CH4 emissions in Alaskan costal lowlands and East Siberian lowlands. Our study also indicates that the precision and coverage of measurements need to be improved to achieve more accurate high-resolution estimates.
