The CO 2 flux on Biebrza Wetlands in Poland is analysed by ground measurements and satellite data. The main goal of this study was to determine the Net Ecosystem Exchange (NEE) for sedges, reeds and grasses using meteorological, soil and vegetation parameters and satellite indicators. The calculations are based on chamber measurements scattered over the main habitats and meteorological registrations, and compared to the 2015–2022 eddy covariance (EC) measurements. The NEE model, to get its distribution throughout the day, was elaborated using modelled gross primary production GPP and modelled respiration RESP. The model of ecosystem respiration was developed applying air temperature, soil temperature, soil moisture and NDVI. The satellite vegetation index NDVI, surface and air difference temperature (T S -T A ), and NDII proved to be good predictors for the daily mean GPP. The data from the EC tower was used to compare the results from modelled NEE and CO 2 flux.
The main goal of this study was the determination of carbon flux model for the area of the Biebrza Wetlands, to evaluate their potential as carbon sinks, based on meteorological, soil and vegetation parameters. Based on ground data and satellite data (images from the Sentinel -1, -2, -3 and MODIS) acquired in 2017–2019, a statistical model was used that combines the carbon dioxide stream, latent heat and soil moisture. The flux of carbon dioxide and latent heat were recorded using the eddy covariance tower, which allows for very high sensitivity measurements at a frequency of 20 Hz. Soil moisture was measured at various points in the Biebrza National Park at a depth of 5 cm with a time resolution of 15 minutes. Later, the field parameters were replaced with their satellite equivalents (e.g. soil moisture - [1]), which made it possible to obtain carbon dioxide flux values for any point within the park. The assessment of carbon fluxes is performed at 1-km spatial scale. Carbon measurements done by the chamber method were used to verify the model. All data were extracted using a R programming language script. Further analyses were carried out in R and Statistica program, the satellite image processing was performed using a ERDAS and program R.