Recent decrease of the impact of tropical temperature on the carbon cycle linked to increased precipitation

Wenmin Zhang, Guy Schurgers, Josep Peñuelas, Rasmus Fensholt, Hui Yang, Jing Tang, Xiaowei Tong, Philippe Ciais, Martin BrandtNature Communications 2023 · DOI

TL;DR

The sensitivity of the global carbon cycle's response to tropical temperature fluctuations has decreased significantly in recent decades, largely due to increased precipitation. This suggests a decoupling of temperature's impact on carbon uptake, driven by a shift towards wetter tropical conditions.

What Problem Does This Paper Address?

The Earth's carbon cycle is a critical component of the climate system, and its variability is strongly influenced by atmospheric CO2 growth rate (CGR). For decades, it's been understood that tropical temperature plays a major role in driving CGR fluctuations. Specifically, the sensitivity of CGR to tropical temperature (denoted as $Y_{CGR}$) was observed to increase significantly from the 1960s onwards. This trend was concerning, as it implied that tropical temperature variations were becoming increasingly potent drivers of atmospheric CO2 changes, potentially exacerbating climate change. However, the trajectory of this sensitivity beyond the early 2000s and the underlying reasons for any changes remained unclear. This paper aims to re-examine this relationship over a longer period and investigate the factors driving any observed trends in $Y_{CGR}$.

Methods

The researchers employed a multi-pronged approach:

  • Long-term CO2 Data Analysis: They utilized long-term atmospheric CO2 records from Mauna Loa and the South Pole, along with climatic data from CRU-4 and ERA5, to calculate the CGR.
  • Sensitivity Calculation ($Y_{CGR}$): They computed $Y_{CGR}$ using a 20-year moving window analysis. This involved multiple regression techniques to isolate the impact of tropical temperature on CGR while accounting for other climate variables like precipitation and solar radiation. They used two regression approaches (M1 and M2) to ensure robustness.
  • Drought and Water Stress Analysis: To understand the role of water availability, they analyzed $Y_{CGR}$ under different dryness conditions, categorized using standardized deviations of precipitation, scPDSI, and SPEI.
  • Dynamic Vegetation Modeling (LPJ-GUESS): A sophisticated dynamic vegetation model was used to disentangle the contributions of various factors (climate variables, CO2 concentration, nitrogen deposition, land cover change) to the observed changes in carbon cycle sensitivity. This involved running simulations where individual drivers were held constant while others varied.
  • Earth System Model (ESM) Intercomparison: They analyzed outputs from 33 Earth System Models (ESMs) participating in the Coupled Model Intercomparison Project Phase 6 (CMIP6) to assess how well these models capture the observed trends and to compare their simulations of tropical carbon sink sensitivity to temperature under varying water conditions.

Datasets

  • Atmospheric CO2 Data: Mauna Loa Observatory (1959-2020) and South Pole station (1980-2020).
  • Climatic Data:
    • CRU-4 (1960-2020): Monthly temperature, precipitation, and solar radiation (cloud cover).
  • ERA5 (1950-present): Fifth-generation ECMWF reanalysis data.
  • GPCC (1960-2019): Monthly gauge-based precipitation data.
  • scPDSI (self-calibrating Palmer Drought Severity Index): For detecting water stress.
  • SPEI (Standardized Precipitation and Evaporation Index): For water availability.
  • GLEAM (v3) (1980-2020): Monthly root-zone soil moisture.
  • GRACE-REC (1979-2018): Reconstructed terrestrial water storage.
  • Land Cover Data: ESA CCI land cover (for excluding water bodies and built-up land).
  • ENSO Data: Multivariate ENSO Index Version 2 (MEI.v2) (1979-2021).
  • CMIP6 Model Outputs: Simulated NBP, temperature, precipitation, and solar radiation (1960-2014) from 33 ESMs.
  • TRENDY Project DGVM Outputs: Monthly NBP data (1960-2019) from 15 dynamic global vegetation models.

Key Insights and Ideas

  • Reversal of a Concerning Trend: The paper reveals that the previously observed increasing trend in the sensitivity of the carbon cycle to tropical temperature ($Y_{CGR}$) has reversed. This is a crucial finding, as it suggests that the carbon cycle might be becoming less vulnerable to temperature-driven CO2 fluctuations.
  • Precipitation as the Dominant Driver: The study strongly implicates increased precipitation in the tropics as the primary factor behind the decreased $Y_{CGR}$. This highlights the critical role of water availability in modulating the temperature-carbon cycle feedback.
  • Decoupling of Temperature and Carbon Cycle: The findings suggest a "decoupling" of the impact of tropical temperature on the carbon cycle. This means that even as temperatures continue to rise, their influence on the rate of CO2 uptake might be diminishing due to wetter conditions.
  • Bi-decadal Oscillations: The variations in $Y_{CGR}$ are significantly correlated with bi-decadal changes in precipitation, indicating a cyclical relationship that influences carbon cycle sensitivity over longer timescales.
  • Model Limitations: While ESMs generally capture the coupling between tropical NBP and temperature, they show considerable variation in their ability to accurately represent the differences in sensitivity under very wet versus very dry conditions. This points to areas where model development is needed.

Key Results

  • $Y_{CGR}$ increased by approximately 200% from 1960-1979 to 1979-2000.
  • However, $Y_{CGR}$ then decreased by approximately 117% from 2001-2020, returning to levels similar to the 1960s.
  • Long-term variations in $Y_{CGR}$ are significantly correlated with changes in precipitation at a bi-decadal scale (correlation coefficient of -0.81).
  • The LPJ-GUESS model simulations indicate that precipitation is the primary driver of interannual variations in modeled NBP sensitivity to temperature.
  • The majority of CMIP6 ESMs (29 out of 33) show a strong coupling between tropical NBP and temperature, but their ability to capture the differences in sensitivity under varying moisture conditions is limited.

Limitations and Future Work

  • Model Uncertainties: The paper acknowledges that even advanced ESMs have limitations in capturing the complex dynamics of the carbon cycle across the tropics, particularly concerning the influence of water availability. This leads to uncertainties in assessing the primary drivers of $Y_{CGR}$.
  • Regional Variations: While the study focuses on tropical regions, the paper notes that regional changes in water availability might contribute to variations in $Y_{CGR}$, suggesting a need for more spatially resolved analyses.
  • Future Research: The authors emphasize the need for further research to better understand the complex interplay of processes regulating the terrestrial carbon cycle's response to drought. Improving the constraint on the spread of models regarding carbon cycle sensitivity to temperature is also highlighted as crucial for reducing uncertainties in future climate predictions.

Summary updated

Abstract

The atmospheric CO<sub>2</sub> growth rate (CGR) variability is largely controlled by tropical temperature fluctuations. The sensitivity of CGR to tropical temperature [Formula: see text] has strongly increased since 1960, but here we show that this trend has ceased. Here, we use the long-term CO<sub>2</sub> records from Mauna Loa and the South Pole to compute CGR, and show that [Formula: see text] increased by 200% from 1960-1979 to 1979-2000 but then decreased by 117% from 1980-2001 to 2001-2020, almost returning back to the level of the 1960s. Variations in [Formula: see text] are significantly correlated with changes in precipitation at a bi-decadal scale. These findings are further corroborated by results from a dynamic vegetation model, collectively suggesting that increases in precipitation control the decreased [Formula: see text] during recent decades. Our results indicate that wetter conditions have led to a decoupling of the impact of the tropical temperature variation on the carbon cycle.
BibTeX
@inproceedings{zhang2023recent,
  author = {Wenmin Zhang and Guy Schurgers and Josep Peñuelas and Rasmus Fensholt and Hui Yang and Jing Tang and Xiaowei Tong and Philippe Ciais and Martin Brandt},
  title = {{Recent decrease of the impact of tropical temperature on the carbon cycle linked to increased precipitation}},
  booktitle = {Nature Communications},
  year = {2023},
  doi = {10.1038/s41467-023-36727-2},
}
Figure