Carbon Cycle Data Assimilation and Inverse modelling

Model calibration and greenhouse gas emissions estimations

We study the interactions between the atmosphere and the earth’s surface focussing on exchanges of carbon, water, and energy at the vegetated land surface. This also includes human interactions and perturbations to the Earth system such as the burning of fossil fuel. We do this by a methodology commonly referred to as model-data fusion.

Our work involves developing and applying simulation models that capture these interactions over various spatial and temporal scales, allowing us to impose robust observational constraints on the carbon, water and energy fluxes. 

An important aspect of our research is on integrating observation with models to obtain a consistent view of the global carbon cycle. We use both remotely sensed observations such as soil moisture, sun-induced fluorescence (SIF), vegetation optical depth (VOD) and XCO2 as well as in-situ ecosystem and atmospheric observations such as those provided by ICOS.

The goal of our research is to improve our process understanding of terrestrial ecosystem functioning. We also contribute to the development of a greenhouse gas Monitoring Verification Support (CO2MVS) capacity to assist emissions reporting to the UN Climate Change Convention Secretariat.

Links: 

ICOS
CO2MVS
UN Climate Change Convention Secretariat

Main activity fields

Regional atmospheric inversions

Regional atmospheric inversions for estimation of CO₂ and CH₄ fluxes and, in collaboration with Combustion Physics, also for aerosol emissions. We use the Lund University Modular Inversion Algorithm (LUMIA) as the main inversion system in our projects, Figure 1 below shows an example output of LUMIA.

Links: 
Combustion Physics at Lund University
Lund University Modular Inversion Algorithm (LUMIA)

Carbon cycle data assimilation

Carbon cycle data assimilation for integrating observations of the land surface and the carbon cycle with process-based models of the terrestrial carbon cycle to improve our understanding of the terrestrial carbon and water cycles. We are one of the main contributors to the Terrestrial Community Carbon Assimilation System (TCCAS) developed within the Land surface Carbon Constellation (LCC) project. We also use the LPJ-GUESS dynamic global vegetation model in a data assimilation setting using EC flux measurements to calibrate model process parameters in collaboration with Mathematical Statistics.

Links: 
Terrestrial Community Carbon Assimilation System (TCCAS)
Land surface Carbon Constellation  project (LCC)
 Mathematical Statistics at Lund University

Fossil fuel data assimilation

Fossil fuel data assimilation for quantifying fossil fuel CO2 emissions globally at high spatial resolution by assimilating in-situ and remotely sensed observations together with socio-economic proxy data into a model of fossil fuel emissions. 

Research projech description in the LU Research Portal

Quantitative Network Design

Quantitative Network Design is methodology closely linked to data assimilation that can be used to evaluate of a set of (candidate) networks in terms of how well they constrain a given target quantity. We use this methodology to evaluate for instance arctic observation networks for constraining permafrost extent or arctic carbon fluxes. Another application is to evaluate the added value of future satellite missions. This work is done in collaboration with the European Space Agency (ESA).
 

Projects

We coordinate / participate in the following projects:

• Arctic-PASSION
   
• AVENGERS
   
• CORSO
   
• ESA-TCCAS 
   
• Greenfeedback
   
• IM4CA