Constraining Uncertainty in Projected Gross Primary Production With Machine Learning

Manuel Schlund,Markus Reichstein,Gustau Camps‐Valls,Veronika Eyring,Pierre Friedlingstein,Pierre Gentine

doi:10.1029/2019jg005619

Abstract

AbstractThe terrestrial biosphere is currently slowing down global warming by absorbing about 30% of human emissions of carbon dioxide (CO2). The largest flux of the terrestrial carbon uptake is gross primary production (GPP) defined as the production of carbohydrates by photosynthesis. Elevated atmospheric CO2 concentration is expected to increase GPP (“CO2 fertilization effect”). However, Earth system models (ESMs) exhibit a large range in simulated GPP projections. In this study, we combine an existing emergent constraint on CO2 fertilization with a machine learning approach to constrain the spatial variations of multimodel GPP projections. In a first step, we use observed changes in the CO2 seasonal cycle at Cape Kumukahi to constrain the global mean GPP at the end of the 21st century (2091–2100) in Representative Concentration Pathway 8.5 simulations with ESMs participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5) to 171 ± 12 Gt C yr−1, compared to the unconstrained model range of 156–247 Gt C yr−1. In a second step, we use a machine learning model to constrain gridded future absolute GPP and gridded fractional GPP change in two independent approaches. For this, observational data are fed into the machine learning algorithm that has been trained on CMIP5 data to learn relationships between present‐day physically relevant diagnostics and the target variable. In a leave‐one‐model‐out cross‐validation approach, the machine learning model shows superior performance to the CMIP5 ensemble mean. Our approach predicts an increased GPP change in northern high latitudes compared to regions closer to the equator.

Highlights

The response of the terrestrial carbon cycle to changes in atmospheric CO2 and climate is a major source of uncertainty in climate projections (Bodman et al, 2013; Booth et al, 2012; M. Collins et al, 2013)
We use observed changes in the CO2 seasonal cycle at Cape Kumukahi to constrain the global mean gross primary production (GPP) at the end of the 21st century (2091–2100) in Representative Concentration Pathway 8.5 simulations with Earth system models (ESMs) participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5) to 171 ± 12 Gt C yr−1, compared to the unconstrained model range of 156–247 Gt C yr−1
In Wenzel, Cox, et al (2016), the observed atmospheric CO2 concentration at Cape Kumukahi, Hawaii (KUM; 19.5°N, 154.8°W) (Keeling et al, 2005) was used to constrain the GPP change in the 1%BGC run resulting from a doubling of atmospheric CO2 concentrations to 32 ± 9% for extratropical ecosystems (30–90°N)

Summary

Introduction

The response of the terrestrial carbon cycle to changes in atmospheric CO2 and climate is a major source of uncertainty in climate projections (Bodman et al, 2013; Booth et al, 2012; M. Collins et al, 2013). The response of the terrestrial carbon cycle to changes in atmospheric CO2 and climate is a major source of uncertainty in climate projections Collins et al, 2013). Future terrestrial carbon sensitivity is mainly driven by two feedback mechanisms: the concentration‐carbon and the climate‐carbon feedbacks Collins et al, 2013; Friedlingstein et al, 2006; Gregory et al, 2009). The first one is connected to the CO2 fertilization effect (Walker et al, 2020) where elevated atmospheric CO2 concentrations increase photosynthesis rates and generally lead to a higher terrestrial carbon uptake, constituting a negative feedback. The second one, the climate‐carbon feedback, is driven by temperature and precipitation changes leading to a smaller land

Methods

Results

Conclusion