Using atmospheric CO2 observations from site level and retrievals from a space-born sensor in soil carbon model evaluation at global scale
Abstract
An accurate simulation of soil carbon stocks and fluxes using global vegetation models is of crucial importance for projections of future climate change, since the soil stores large amounts of carbon and a large release of this carbon to the atmosphere would lead to strong climatic feedbacks.
In this work we are using atmospheric CO2 observations to evaluate two different soil carbon models implemented in the global biosphere model JSBACH. We have performed global JSBACH simulations using the two different soil carbon models, CBALANCE (CBA) and YASSO (YAS). CBA is the original soil carbon model of the JSBACH. It is dependent of soil temperature with a Q10 formulation and has a linear soil moisture dependency. YAS is a newer model that has been calibrated against several observations. It has a Gaussian dependency on air temperature and exponential dependency on precipitation. The simulations with JSBACH started from 1860. Interactive fire model was used and land use change and land cover change were calculated dynamically. The gross primary production was identical between the two different soil carbon model simulations. Atmospheric CO2 mole fractions for 2001-2014 were simulated with TM5 transport model by using the biospheric fluxes from JSBACH together with ocean and anthropogenic carbon fluxes from the CarbonTrackerEurope. These simulated concentrations were then used in evaluation against calibrated in-situ measurements and spaceborn retrievals from the GOSAT. The YAS was better at capturing the observed seasonal cycle amplitude (SCA) in northern latitudes than CBA. At Pallas in the Finnish Lapland the SCA predicted by YAS was 1% less than the observed, but CBA overestimated the SCA by 44%. However, in the lower latitudes YAS had too low SCA and CBA was instead performing better. At Mauna Loa YAS estimated the SCA to be only 51% of the observed, whereas CBA estimated it to be 87% of the observed. The comparison against GOSAT results showed similar behaviour in larger regions. The reason for these differences originated from the tropical zone, where the annual cycle of the heterotrophic respiration was driven by moisture conditions. The YAS predicted much earlier increase to the higher summertime values in the year, than CBA. These differences in the tropics caused differences in the CO2 dynamics globally.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.B23B..07T
- Keywords:
-
- 0428 Carbon cycling;
- BIOGEOSCIENCESDE: 0439 Ecosystems;
- structure and dynamics;
- BIOGEOSCIENCESDE: 0480 Remote sensing;
- BIOGEOSCIENCESDE: 1615 Biogeochemical cycles;
- processes;
- and modeling;
- GLOBAL CHANGE