WP1

WP1 Natural emission modeling (Lead WSL)

Objectives of this activity

  1. To understand the observed variability in atmospheric methane concentrations in the recent past, including an integrated global and regional analysis of the main effects of extreme weather events (e.g. ENSO) on methane source emissions and sinks (link to WP1 and WP3)
  2. To understand the impact of model differences and different land cover and climate inputs on interannual variability in methane emissions (link to WP1, WP3, and external partners)

Approach and Methods:
We will use the LPJ-wslCH4 dynamic global vegetation model (DGVM), which has been adapted during MAIOLICA-I at WSL to model spatially and temporally resolved wetland, rice, and fire CH4 emissions. LPJ-wslCH4 is a version of the internationally well established community DGVM LPJ which includes improved hydrological dynamics and can be applied to any spatial resolution (Sitch et al., 2003; Gerten et al., 2004). Through a collaboration between WSL and EPFL during MAIOLICA-I, we developed a wetland module specifically to study interannual variability in global and regional wetland CH4 emissions with a model parameterization which not only accounts for variations in available biomass, as is common with wetland process modeling, but also wetland area over multiple decades. Additionally, the existing fire module in LPJ-wslCH4 was recently adapted at WSL to model fire emissions for a number of trace gases. Thus, LPJ-wslCH4 is able to estimate interannual variability in CH4 emissions from three major land surface processes (i.e. natural wetlands, rice, and fire), which together are estimated to comprise over 80% of total interannual variability in CH4 emissions globally (Bousquet et al., 2006).

Main tasks of this work package:
Task 1: Preparation of recent past model runs. We will provide emissions from biomass burning, wetlands and rice to both WP2 and WP3 as input for their first model simulations for the recent past (1990-2010). Additionally, we will extend the existing data sets out to 2010 and fill in partly missing periods in the early 1990s. We will develop a merged data set for wetlands and rice covering the entire 1990-2010 time period for use in WP2 and WP3.

Task 2: Sensitivity study. In parallel with the recent past simulations in WP2 and WP3, we will perform a sensitivity study with the LPJwslCH4 model, in which we vary the land cover model input, climate model input, and the model itself. We will use the runs from MAIOLICA-I to test the sensitivity of the model to land cover input. To test the sensitivity of the model to climate inputs, we will perform a factorial study in which we allow only one climate input (precipitation, temperature, cloud cover, CO2 concentrations) at a time to vary interannually. To test the sensitivity of CH4 emissions and interannual variations in CH4 emissions to different process models, we will work closely with the Wetland and Wetland Methane Model Intercomparison of Models Project (WETCHIMP) to provide wetland emission data sets for the years 1994-2004 from additional wetland models (e.g. ORCHIDEE, CLMCN, ARVE, etc). The runs developed and analyzed in this task will be used in combination with an Eulerian-based atmospheric transport model to test the influence of climate and land cover variations on the accumulation of CH4 in the atmosphere (CH4 growth rate). This will be done in collaboration with M. Rigby at the University of Bristol within the international Advanced Global Atmospheric Gases Experiment (AGAGE).

Task 3: Comparison and validation. We will compare our runs from Tasks 1 and 2 with the results of the first inverse modeling task in WP2. This will allow us to separate effects by emission category, region, and year, and to provide a better understanding of what processes or inversion techniques need to be improved to reduce uncertainty in estimates of biogenic CH4 emissions and in estimates of CH4 interannual variability (see Task 5 in WP2). In addition, we will work with S. Seneviratne (ETHZ) to validate the hydrology dynamics in LPJ-wslCH4 with satellite hydrology data.

Task 4: Preparation of second reference runs for recent past and future CH4 emissions. We will use the validated CH4 emissions and the improvements made to LPJ-wslCH4 during Task 3 to develop a second set of reference runs for the recent past to be used in a second reference simulation in WP2 and WP3. Furthermore, we will use the model improvements from Task 3, in combination with our published technique (Hodson et al., 2011) for hindcasting satellite-derived wetland area, to forecast CH4 emissions from wetlands and possibly also biomass burning for the 21st century. These forecasted interannually varying CH4 emissions will be delivered to WP3 for their future scenario model runs.

Task 5: Integrated ENSO analysis. In collaboration with WP2 and WP3, we will build upon work published in MAIOLICA-I in which we studied the interaction of wetland emissions with extreme wet and dry events corresponding to the El-Nino Southern Oscillation (ENSO) (Hodson et al., 2011). With the addition of the fire module into LPJ-wsl, the ability to study the link between ENSO and the main CH4 sink, the OH radical, in WP3 and CH4 emissions from the inverse modeling results in WP2, we will perform the first integrated assessment of the global and regional effect of ENSO on all of the major sources and sinks affected by climate (wetland, rice, and fire emissions and OH sink).

Deliverables and expected results:

  1. Gridded global wetland and fire methane emissions for 1990-2010.
  2. Gridded global wetland methane emissions for the 21st century
  3. Analysis of the major factors (climate, land cover, etc.) driving variations in terrestrial CH4 emissions
  4. Improved understanding of the impact of ENSO on CH4 variability
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License