Soil moisture–atmosphere feedback dominates land carbon uptake variability

Vincent Humphrey, Alexis Berg, Philippe Ciais, Pierre Gentine, Martin Jung, Markus Reichstein, Sonia I. Seneviratne, Christian FrankenbergNature 2021 · DOI

TL;DR

This study reveals that year-to-year fluctuations in global land carbon uptake are overwhelmingly driven by soil moisture variability, accounting for 90% of this inter-annual variability. This influence is primarily indirect, mediated through soil moisture's impact on temperature and atmospheric dryness (VPD), which in turn affect photosynthesis.

What Problem Does This Paper Address?

The terrestrial biosphere plays a critical role in regulating atmospheric CO2 concentrations by absorbing carbon. However, the year-to-year variability in this "land carbon uptake" is substantial and its primary drivers have been a subject of debate. While temperature and water availability are known to influence carbon uptake, it's been unclear whether temperature or soil moisture is the dominant factor at a global scale. Existing climate models show a strong influence of soil moisture, but this hasn't always been apparent from land surface models or observational analyses, leading to conflicting perspectives. This paper aims to reconcile these views by disentangling the direct and indirect effects of soil moisture on land carbon uptake.

Methods

The researchers employed factorial climate model simulations from the Global Land-Atmosphere Coupling Experiment (GLACE-CMIP5). The core of their method involved comparing two simulation setups: a "control" simulation (CTL) that included natural soil moisture variability, and an "experiment A" where soil moisture variability was artificially removed, forcing it to follow a prescribed seasonal cycle. By comparing the carbon uptake (Net Biome Production, NBP) anomalies between these two simulations, they could isolate the impact of soil moisture variability. They further decomposed these impacts into "direct" soil moisture effects and "indirect" effects mediated through land-atmosphere coupling (LAC), which amplifies temperature and humidity anomalies. A sensitivity analysis using linear regression was also performed to attribute NBP variability to different meteorological drivers.

Datasets

The study utilized outputs from four Earth System Models (ESMs) participating in the GLACE-CMIP5 experiment: CCSM4, ECHAM6, GFDL, and IPSL. These models provided simulated data for carbon fluxes (NBP, GPP, respiration), soil moisture, temperature, and vapor pressure deficit (VPD). The study also referenced observational datasets for validation, including FLUXCOM RS+METEO, CAMS atmospheric CO2 inversion, ERA5-Land, ESA CCI Soil Moisture, CRU TS, and GSWP3.

Key Insights and Ideas

  • Soil Moisture is King: The most striking finding is that soil moisture variability, not temperature or VPD directly, drives the vast majority (90%) of year-to-year global land carbon uptake fluctuations.
  • Indirect Effects Dominate: The primary mechanism is indirect: soil moisture influences temperature and VPD through land-atmosphere coupling (LAC), and these altered atmospheric conditions then impact photosynthesis and carbon uptake.
  • Reconciling Debates: This work bridges the gap between models showing strong soil moisture influence and observational studies that sometimes highlight temperature or VPD. The indirect pathway explains why temperature and VPD appear important – they are themselves responding to soil moisture.
  • LAC Hotspots: Regions with strong land-atmosphere coupling (often semi-arid and tropical areas) are also the areas where carbon uptake variability is most pronounced.
  • Model Behavior: The study highlights that while models agree on the dominant role of soil moisture, the exact partitioning between direct and indirect effects might vary depending on the specific model's parameterization of land-atmosphere interactions.

Key Results

  • Suppressing non-seasonal soil moisture variability reduced the inter-annual variability (IAV) in global mean Net Biome Production (NBP) by 91%.
  • Indirect effects, mediated by land-atmosphere coupling (LAC), accounted for approximately 60% of the global NBP IAV.
  • Direct soil moisture effects contributed about 20% to the global NBP IAV.
  • Temperature and VPD effects independent of soil moisture (non-LAC) and radiation effects accounted for the remaining 9% and 11%, respectively.

Limitations and Future Work

  • Model Dependence: While models agree on the overall dominance of soil moisture, the precise attribution of NBP variability to direct vs. indirect effects can be model-dependent due to differing parameterizations of land-atmosphere interactions.
  • Observational Constraints: Directly replicating the factorial experiment with observations is not feasible, making it challenging to fully validate the findings. However, the study notes good agreement between model sensitivities and observational estimates for NBP IAV.
  • Future Projections: The authors suggest that improved physical and holistic representations of vegetation responses to soil and atmospheric dryness are crucial for reducing uncertainties in future terrestrial carbon flux projections. Understanding how LAC and direct soil moisture effects might respond differently to climate change is also highlighted as an area for future research.

Summary updated

Abstract

Year-to-year changes in carbon uptake by terrestrial ecosystems have an essential role in determining atmospheric carbon dioxide concentrations<sup>1</sup>. It remains uncertain to what extent temperature and water availability can explain these variations at the global scale<sup>2-5</sup>. Here we use factorial climate model simulations<sup>6</sup> and show that variability in soil moisture drives 90 per cent of the inter-annual variability in global land carbon uptake, mainly through its impact on photosynthesis. We find that most of this ecosystem response occurs indirectly as soil moisture-atmosphere feedback amplifies temperature and humidity anomalies and enhances the direct effects of soil water stress. The strength of this feedback mechanism explains why coupled climate models indicate that soil moisture has a dominant role<sup>4</sup>, which is not readily apparent from land surface model simulations and observational analyses<sup>2,5</sup>. These findings highlight the need to account for feedback between soil and atmospheric dryness when estimating the response of the carbon cycle to climatic change globally<sup>5,7</sup>, as well as when conducting field-scale investigations of the response of the ecosystem to droughts<sup>8,9</sup>. Our results show that most of the global variability in modelled land carbon uptake is driven by temperature and vapour pressure deficit effects that are controlled by soil moisture.
BibTeX
@inproceedings{humphrey2021soil,
  author = {Vincent Humphrey and Alexis Berg and Philippe Ciais and Pierre Gentine and Martin Jung and Markus Reichstein and Sonia I. Seneviratne and Christian Frankenberg},
  title = {{Soil moisture–atmosphere feedback dominates land carbon uptake variability}},
  booktitle = {Nature},
  year = {2021},
  doi = {10.1038/s41586-021-03325-5},
}
Figure