Large-scale assessment of gross primary productivity on different peatland types using a satellite-based time series

Abstract

Peatlands store large amounts of terrestrial carbon and any changes to their carbon balance could cause large changes in the greenhouse gas balance of the Earth's atmosphere. There is still much uncertainty about how the dynamics of peatlands are affected by climate and land use change. Current field-based methods of estimating annual carbon exchange between peatlands and the atmosphere include flux chambers and eddy covariance towers. However, remote sensing has several advantages over these traditional approaches in terms of cost and spatial coverage. In this study we produced a time series from 1998 to 2016 consisting of annual gross and net primary productivity using the satellite-based Biosphere Energy Transfer Hydrology (BETHY/DLR) model for the entire federal state Mecklenburg-Western Pomerania (MV, Germany). All landscape types except peatlands were excluded and the remaining pixels were ordered by their main peatland type. The four main types were percolation peatlands (41 %), terrestrialisation peatlands (28 %), flood peatlands (14 %), and paludification peatlands (13 %). All together they cover 95 % of the area of peatlands in MV. Regarding gross primary productivity, the results show significant differences between these four peatland types. Percolation mires and paludification mires have the higest amount of Carbon sequestration followed by the terrestrialisation peatlands. At some distance flood peatlands (including coastal and floodplain peatlands) seemed to sequestrate much lower amount of carbon. A certain shape within the time series, which is clearly regognizable in all four peatland types, was compared to climate parameters. Preliminary results were also produced considering the different grades of drainage (not, moderatly, strongly and extremely drained) for all types but these results have to be interpreted very carefully since rewetting processes mabye started during the period between 1998 and 2016.