Search

Soil respiration is the second largest flux in the global carbon cycle, yet the underlying below-ground process, carbon dioxide (CO2) production, is not well understood because it can not be measured in the field. CO2 production has frequently been calculated from the vertical CO2 diffusive flux divergence, known as "soil-CO2 profile method". This relatively simple model requires knowledge of soil CO2 concentration profiles and soil diffusive properties. Application of the method for a tropical lowland forest soil in Panama gave inconsistent results when using diffusion coefficients (D) calculated based on relationships with soil porosity and moisture ("physically modeled" D). Our objective was to investigate whether these inconsistencies were related to (1) the applied interpolation and solution methods and/or (2) uncertainties in the physically modeled profile of D. First, we show that the calculated CO2 production strongly depends on the function used to interpolate between measured CO2 concentrations. Secondly, using an inverse analysis of the soil-CO2 profile method, we deduce which D would be required to explain the observed CO2 concentrations, assuming the model perception is valid. In the top soil, this inversely modeled D closely resembled the physically modeled D. In the deep soil, however, the inversely modeled D increased sharply while the physically modeled D did not. When imposing a constraint during the fit parameter optimization, a solution could be found where this deviation between the physically and inversely modeled D disappeared. A radon (Rn) mass balance model, in which diffusion was calculated based on the physically modeled or constrained inversely modeled D, simulated observed Rn profiles reasonably well. However, the CO2 concentrations which corresponded to the constrained inversely modeled D were too small compared to the measurements. We suggest that, in well-structured soils, a missing description of steady state CO2 exchange fluxes across water-filled pores causes the soil-CO2 profile method to fail. These fluxes are driven by the different diffusivities in inter- vs. intra-aggregate pores which create permanent CO2 gradients if separated by a "diffusive water barrier". These results corroborate other studies which have shown that the theory to treat gas diffusion as homogeneous process, a precondition for use of the soil-CO2 profile method, is inaccurate for pore networks which exhibit spatial separation between CO2 production and diffusion out of the soil.

Sunlight induces photochemical mineralisation of chromophoric dissolved organic matter (CDOM) to dissolved inorganic carbon (DIC) in inland waters, resulting in carbon dioxide (CO2) emissions to the atmosphere. Photochemical rate modelling is used to determine sunlight-induced CO2 emissions on large spatial and temporal scales. A sensitive model parameter is the wavelength-specific photochemical CDOM reactivity, the apparent quantum yield (AQY). The modelling studies so far assume that AQY spectra determined for single lakes and on single occasions represent larger spatial and temporal scales. Here, we studied a humic boreal lake in Sweden. We measured AQY spectra for photochemical DIC production monthly between June and November 2014 and parameterised a photochemical rate model. Photochemical reactivity increased slightly during the open water period, likely due to a high rainfall event with consecutive mixing in autumn that increased availability of highly photoreactive CDOM. However, the variability in AQY spectra over time was much smaller than previously reported variability in AQY spectra across lakes. Yet, using either the AQY spectrum from the least or from the most photoreactive water sample resulted in a 5-fold difference in simulated annual DIC photoproduction (2.0 ± 0.1 and 10.3 ± 0.7 g C m−2 yr−1, respectively). Using the monthly measured AQY spectrum to simulate DIC photoproduction for month-long time periods resulted in an apparent time lag between irradiance and DIC photoproduction. This suggested that temporal variability in AQY spectra occurs on shorter time scales. Therefore, we parameterised the model with the pooled AQY spectrum of six monthly measurements. Simulated DIC photoproduction for three years (2012–2014) averaged 4.5 ± 0.2 g C m−2 yr−1, which represented 3 % of the mean CO2 emissions from this lake. We conclude that (1) it may be recommendable to conduct repeated AQY measurements across the season for more accurate simulation of annual photochemical DIC production in lakes and (2), in agreement with previous studies, direct CDOM photomineralisation makes only a minor contribution to mean CO2 emissions from Swedish humic lakes.