 |
| Titel |
Global distribution of carbon turnover times in terrestrial ecosystems |
| VerfasserIn |
Nuno Carvalhais, Matthias Forkel, Myroslava Khomik, Jessica Bellarby, Martin Jung, Mirco Migliavacca, Mingquan Mu, Sassan Saatchi, Maurizio Santoro, Martin Thurner, Ulrich Weber, Bernhard Ahrens, Christian Beer, Alessandro Cescatti, James T. Randerson, Markus Reichstein |
| Konferenz |
EGU General Assembly 2015
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 17 (2015) |
| Datensatznummer |
250113529
|
| Publikation (Nr.) |
 EGU/EGU2015-13739.pdf |
|
|
|
|
|
| Zusammenfassung |
| The response of the carbon cycle in terrestrial ecosystems to climate variability remains one
of the largest uncertainties affecting future projections of climate change. This feedback
between the terrestrial carbon cycle and climate is partly determined by the response of
carbon uptake and by changes in the residence time of carbon in land ecosystems, which
depend on climate, soil, and vegetation type. Thus, it is of foremost importance to quantify
the turnover times of carbon in terrestrial ecosystems and its spatial co-variability with
climate.
Here, we develop a global, spatially explicit and observation-based assessment of
whole-ecosystem carbon turnover times (Ï) to investigate its co-variation with climate at
global scale. Assuming a balance between uptake (gross primary production, GPP) and
emission fluxes, Ï can be defined as the ratio between the total stock (C_total) and the output
or input fluxes (GPP). The estimation of vegetation (C_veg) stocks relies on new remote
sensing-based estimates from Saatchi et al (2011) and Thurner et al (2014), while soil carbon
stocks (C_soil) are estimated based on state of the art global (Harmonized World Soil
Database) and regional (Northern Circumpolar Soil Carbon Database) datasets. The uptake
flux estimates are based on global observation-based fields of GPP (Jung et al.,
2011).
Globally, we find an overall mean global carbon turnover time of 23-4+7 years (95%
confidence interval). A strong spatial variability globally is also observed, from
shorter residence times in equatorial regions to longer periods at latitudes north of
75ºN (mean Ï of 15 and 255 years, respectively). The observed latitudinal pattern
reflect the clear dependencies on temperature, showing increases from the equator
to the poles, which is consistent with our current understanding of temperature
controls on ecosystem dynamics. However, long turnover times are also observed in
semi-arid and forest-herbaceous transition regions. Furthermore, based on a local
correlation analysis, our results reveal a similarly strong association between Ï and
precipitation.
A further analysis of carbon turnover times as simulated by state-of-the-art coupled
climate carbon-cycle models from the CMIP5 experiments reveals wide variations between
models and a tendency to underestimate the global Ï by 36%. The latitudinal patterns
correlate significantly with the observation-based patterns. However, the models show
stronger associations between Ï and temperature than the observation-based estimates. In
general, the stronger relationship between Ï and precipitation is not reproduced
and the modeled turnover times are significantly faster in many semi-arid regions.
Ultimately, these results suggest a strong role of the hydrological cycle in the carbon
cycle-climate interactions, which is not currently reproduced by Earth system models. |
|
|
|
|
|
|