|
Titel |
Photoproduction of dissolved inorganic carbon in Swedish lakes |
VerfasserIn |
B. Koehler, T. Landelius, L. J. Tranvik |
Konferenz |
EGU General Assembly 2012
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 14 (2012) |
Datensatznummer |
250063020
|
|
|
|
Zusammenfassung |
A substantial fraction of the dissolved organic carbon (DOC) in inland waters is mineralized
to dissolved inorganic carbon (DIC) during passage towards the sea. Both microbial and
photochemical mineralization have a share but there is currently no landscape-scale estimate
of the contribution of photomineralization to total lake carbon dioxide emissions, restricting
our understanding of inland-water C cycling. We use 1) DOC absorbance spectra measured
during autumn 2009 in water samples from 1074 lakes distributed across Sweden,
2) light attenuation coefficients estimated based on correlations with absorption
coefficients as established from literature data, 3) cloud-corrected, below-water-surface
downwelling scalar irradiance spectra derived by modeling radiative transfer in the
atmosphere and transmission into the water and 4) an apparent quantum yield spectrum
determined in a humic lake, to calculate spectra of DIC photoproduction from 280 to 600
nm and from the water surface down to the mean lake depths. For each lake, we
calculate DIC photoproduction rates on a daily base and integrate to obtain yearly flux
estimates. Preliminary model results calculated for July 2009 show that, even though
water color differed largely (25%- and 75%-quantiles of specific UV absorption
coefficients at 254 nm (SUV A254) of 6.4 and 9.6 L mg C-1 m-1, respectively),
depth-integrated DIC photoproduction rates showed a relatively small variation
with a 25%-quantile of 12.0 and a 75%-quantile of 13.1 mg C m-2 day-1. These
rather similar DIC photoproduction rates are explained by their different depth
distributions: The brownest lake with a SUV A254 of 12.9 L mg C-1 m-1 had large
surface DIC photoproduction rates of 887.9 mg C m-3 day-1 but photons were
quickly attenuated with depth, with DIC photoproduction rates falling below 1 mg C
m-3 day-1 already at Ë 0.2 m depth (depth-integrated rate of 14.2 mg C m-2
day-1). The clearest lake with a SUV A254 of 1.4 L mg C-1 m-1 had nearly
100-fold smaller surface DIC production rates of 9.4 mg C m-3 day-1 but rates still
reached Ë 1 mg C m-3 day-1 at 2 m depth (depth-integrated rate of 8.4 mg C
m-2 day-1). Across lakes, DIC photoproduction rates correlated positively with
specific absorption coefficients at 420 nm (SA420) (y = 10.3 + 3.1 * x, R2 = 0.76, P <
0.0001). Using this relationship we predict DIC photoproduction rates for a larger data
set of 3853 Swedish lakes for which SA420 has been determined in further recent
sampling campaigns. Assuming that the combined data set (n = 4927) is representative
for the total population of Swedish lakes, we upscale the flux by multiplying the
mean DIC photoproduction rate with the overall lake area to obtain a country-wide,
annual DIC flux estimate. DIC-fluxes from photomineralization are compared to
total carbon dioxide emissions from Swedish lakes available from earlier studies. |
|
|
|
|
|