Vegetation productivity is highly sensitive to temperature and can decline once temperature exceeds an optimum level (Topt). Accurate representation of Topt in Earth System Models (ESMs) is therefore essential for projecting land carbon uptakes and associated climate feedbacks. Prof. NIU Shuli's team at Institute of Geographic Sciences and Natural Resources Research of Chinese Academy of Sciences found that current ESMs systematically overestimate Topt and fail to capture its increase over the past four decades. Further studies revealed that this overestimation was linked to misrepresented water limitation and vegetation structures under high temperatures, providing key pathways for improved projections of future carbon-climate feedbacks. This work was published in One  Earth.

Terrestrial ecosystems play a crucial role in mitigating climate change by absorbing carbon dioxide from the atmosphere. Scientists and policymakers rely on ESMs to estimate ecosystem carbon uptake, which underpins climate policy, carbon budgeting, and land management strategies. The rate of carbon uptake, known as gross primary productivity (GPP), is highly sensitive to temperature and declines rapidly once it exceeds Topt. However, it remains unclear whether ESMs accurately capture Topt, adding uncertainty to land-atmosphere feedbacks and limiting the reliability of long-term climate projections.

Under the guidance of Prof. NIU Shuli, and with the support of Prof. HUANG Yuanyuan and XIA Jianyang, Dr. WANG Yiheng found that ESMs overestimated or failed to capture Topt across 60.3% of ecosystems, especially in arid regions, due to misrepresented water limitations and vegetation structure changes under high temperatures. These biases suggest that models may overestimate the positive effects of warming on land carbon uptake, amplifying carbon-climate feedbacks in a warming climate. 

Moreover, ESMs underestimated the observed temporal increase in Topt, highlighting an urgent need to better understand the acclimation of land ecosystems to rising temperatures. By bridging Topt bias and future GPP projections, this study identifies critical pathways to improve land carbon uptake predictions. Such improvements are essential not only for reliable climate projections, but also for sustaining ecosystem functions and services, and supporting international efforts to mitigate climate change.

Figure: Comparison of modeled and observed Topt (Image by Prof. NIU's team)