Abstract. There is growing interest in developing spatially resolved methane
(CH4) isotopic source signatures to aid in geographic source
attribution of CH4 emissions. CH4 hydrogen isotope measurements
(δ2H–CH4) have the potential to be a powerful tool for
geographic differentiation of CH4 emissions from freshwater
environments, as well as other microbial sources. This is because microbial
δ2H–CH4 values are partially dependent on the δ2H of environmental water (δ2H–H2O), which exhibits
large and well-characterized spatial variability globally. We have refined
the existing global relationship between δ2H–CH4 and
δ2H–H2O by compiling a more extensive global dataset of
δ2H–CH4 from freshwater environments, including wetlands,
inland waters, and rice paddies, comprising a total of 129 different sites,
and compared these with measurements and estimates of δ2H–H2O, as well as δ13C-CH4 and δ13C–CO2 measurements. We found that estimates of δ2H–H2O explain approximately 42 % of the observed variation in
δ2H–CH4, with a flatter slope than observed in previous
studies. The inferred global δ2H–CH4 vs. δ2H–H2O regression relationship is not sensitive to using either
modelled precipitation δ2H or measured δ2H–H2O as the predictor variable. The slope of the global
freshwater relationship between δ2H–CH4 and δ2H–H2O is similar to observations from incubation experiments
but is different from pure culture experiments. This result is consistent
with previous suggestions that variation in the δ2H of acetate,
controlled by environmental δ2H–H2O, is important in
determining variation in δ2H–CH4. The relationship between
δ2H–CH4 and δ2H–H2O leads to significant
differences in the distribution of freshwater δ2H–CH4
between the northern high latitudes (60–90∘ N), relative to
other global regions. We estimate a flux-weighted global freshwater δ2H–CH4 of −310 ± 15 ‰, which is higher
than most previous estimates. Comparison with δ13C measurements
of both CH4 and CO2 implies that residual δ2H–CH4 variation is the result of complex interactions between
CH4 oxidation, variation in the dominant pathway of methanogenesis, and
potentially other biogeochemical variables. We observe a significantly
greater distribution of δ2H–CH4 values, corrected for
δ2H–H2O, in inland waters relative to wetlands, and
suggest this difference is caused by more prevalent CH4 oxidation in
inland waters. We used the expanded freshwater CH4 isotopic dataset to
calculate a bottom-up estimate of global source δ2H–CH4
and δ13C-CH4 that includes spatially resolved isotopic
signatures for freshwater CH4 sources. Our bottom-up global source
δ2H–CH4 estimate (−278 ± 15 ‰) is
higher than a previous estimate using a similar approach, as a result of the
more enriched global freshwater δ2H–CH4 signature derived
from our dataset. However, it is in agreement with top-down estimates of
global source δ2H–CH4 based on atmospheric measurements
and estimated atmospheric sink fractionations. In contrast our bottom-up
global source δ13C-CH4 estimate is lower than top-down
estimates, partly as a result of a lack of δ13C-CH4 data
from C4-plant-dominated ecosystems. In general, we find there is a
particular need for more data to constrain isotopic signatures for
low-latitude microbial CH4 sources.
