Analysis of climate feedback PArameter Changes from Heat Evolution (APACHE)

Effective Climate sensitivity: a study of climate feedback parameter’s time variation

In the context of climate change and the rise of greenhouse gases concentration in the atmosphere, the concept of climate sensitivity emerges as a crucial quantitative measure of the warming associated with increased atmospheric GHG concentration. Understanding climate sensitivity is essential for studying current and future climate changes. The key component in this understanding is the climate feedback parameter, denoted as λ, which describes the Earth’s radiative response to changes in global average 2-metre air temperature. Equilibrium Climate Sensitivity (ECS) represents the Earth’s global average surface temperature reached after an instantaneous doubling of atmospheric CO₂ concentration compared to pre-industrial levels. The relationship between ECS and the λ parameter is direct: a less negative λ value indicates a weaker radiative response of the Earth to a given change in 2-metre air temperature meaning the Earth radiates less energy towards space when it warms leading to greater warming for the same amount of GHG concentration in the atmosphere and thus a greater climate sensitivity to GHG concentrations. Defining and understanding the climate feedback parameter λ is fundamental for accurately assessing ECS and the broader implications for climate change. Currently λ is poorly constrained and subject to significant uncertainties meaning the ECS, which depends on λ, is also highly uncertain. Reducing the uncertainty in ECS estimates involves better constraining λ. This is challenging because λ cannot be directly measured and must therefore be estimated through approaches based on the global energy balance (Otto et al. 2014, Sherwood et al. 2020), which is defined as ΔEEI = ΔF + λΔT, with EEI the Earth Energy Imbalance, F the radiative forcing and T the global mean surface temperature.

From changes in radiative forcing, temperature and earth energy imbalance to climate feedback parameter 

To estimate the λ parameter through the global energy balance equation, variations in radiative forcings, temperatures and planetary heat uptake relative to a defined reference state are quantified. To address uncertainties about the reference state, a method of linearly regressing the radiative response EEI(t) - F(t) against temperature T(t) is used, as in Barnes and Barnes (2015), Chenal et al. (2022), Forster and Gregory (2006) and Gregory et al. (2019).

Climate models suggest that λ varies over time. Recently, a regression approach provided the first observational estimate of temporal variations of λ since 1970 (Meyssignac et al., 2023). In this approach, based on global energy balance, ΔEEI is regressed on ΔF and ΔT for any periods longer than 25 years since 1957, estimating λ variations between 1970 and 2005. This approach uses recent observational data to study climate changes, avoiding reliance on poorly known historical reference states. Additionally, λ has been estimated using three other EEI datasets over more recent periods (24 to 35 years), including the MOHeaCAN EEI product series.

Correlation between climate feedback parameter time-variation and the Pacific Decadal Oscillation 

This study allows for the exploration of temporal variations of λ before the 2000s. λ increased sharply in the mid-1970s until the 1990s, and has been decreasing continuously since then meaning the climate system is less sensitive to GHG concentrations today than in the 1970s .

The correlations between globally gridded SST and the λ time series highlight a significant correlation with the PDO in the Pacific, clearly marked over the period 1970-2005 and still visible over the period 1980-2005. Only the correlation with the PDO is significant, indicating that the multidecadal variations of λ are synchronous with the PDO. This leads to the conclusion that the temporal variations of λ are closely linked to the PDO, not only during the hiatus period, but also at multidecadal time scales.