Matthews, Elaine: NASA ARC (Project Lead)
Project Funding:
2016 - 2020
NRA: 2016 NASA: Interdisciplinary Research in Earth Science
Funded by NASA
Abstract:
Natural wetlands, lakes and reservoirs (WLR) produce the powerful greenhouse gas methane (CH4). Natural wetlands account for about 30% of annual emissions and dominate interannual variations in atmospheric methane growth rates. Lakes and reservoirs also produce CH4, but their emissions have never been comprehensively quantified.
Results from 30+ wetland-CH4 models reveal extensive differences in magnitude, seasonality and geography of simulated emissions that stymie robust conclusions about the role of wetlands in large-scale patterns, trends and interannual variations in atmospheric CH4. However the overwhelmingly contributor to these divergent emissions is differences in wetland representation. Many models now define wetlands (or CH4-producing areas (MPAs)) using monthly inundation data that include, but do not distinguish among, flooded wetlands, lakes, rivers, and irrigated rice. Under- and over-estimates of MPAs are predictable, and the approach fails to capture non-flooded wetlands dominating in high latitudes. In fact, increasingly complex wetland-CH4 models incorporate little wetland-specific information, and use only a small fraction of the hundreds of published CH4-flux observations to develop and evaluate models, in part because there is no link between wetland ecosystems in which fluxes have been measured and the global distribution of those ecosystems.
In response to Interdisciplinary Science subelement 1: Understanding the global sources and sinks of methane, and the importance of improving characterization and representation of wetlands (lakes and reservoirs) to better quantify their role in the contemporary CH4 cycle, we propose the following research.
Task 1. New foundational source data sets, flux syntheses and CH4-centric source classification.
1.1 Develop unique foundational, global gridded data sets of WLR, heavily augmented with CH4-relevant environmental, climatic, and biogeophysical characteristics. 1.2 Synthesize all published WLR flux observations and site characteristics to produce
FLUX+ data sets to support developing and assess ecosystem-specific CH4-flux models. 1.3 Develop CH4-centric classification systems for WLRs, using characteristics identified in Task 1.1, and classify the ecosystems represented in the FLUX+ data set as well as the global distribution of those same ecosystems. This simple step directly links, for the first time, every ecosystem flux observation to the global distribution of those ecosystem classes.
Task 2. Regression and biogeochemical modeling of methane emissions
2.1 Explore observation-anchored multiple-regression models for ecosystem classes identified in the WLRs (Task 1.3) using the observational WLR FLUX+ flux and environmental data to identify the most important factors influencing CH4 emissions in each ecosystem; applying the ecosystem-specific models to the global distribution of WLR ecosystems will calculate monthly methane emissions for WLR for 1985-2015. We anticipate multiple iterations for this effort informed by top-down CarbonTracker-CH4 assessments of emissions. 2.2 Expand the domain of a simple biogeochemical model, developed for boreal wetlands, to the globe using FLUX+wetland and knowledge from developing the ecosystem regression models; run global simulations for all wetland classes for 1985-2015.
Task 3. Top-down evaluation of emissions and models using CarbonTracker-CH4
Conduct simulations constrained with surface, tower, airborne and satellite observations.
3.1 Evaluate every global WLR emission estimate (Task 2) with CarbonTracker-CH4 (CT) to assess realism of new emissions and identify causes of poor performance iterate model development. 3.2 Evaluate every North American WLR emission estimate with high-resolution CarbonTracker-Lagrange (CT-L) as in 3.1 and to compare CT and CT-L results to quantify errors in the coarser-resolution CT.
More details may be found in the following project profile(s):