Breakthrough in estimating fossil fuel CO₂ emissions

A team of scientists led by the University of East Anglia (UEA) has made a major breakthrough in detecting changes in fossil fuel carbon dioxide emissions more quickly and frequently.

The English version of this press release (here) is provided by the University of East Anglia.

In a study published today they quantified regional fossil fuel CO₂ emissions reductions during the Covid-19 lockdowns of 2020-2021, using atmospheric measurements of CO₂ and oxygen (O₂) from the Weybourne Atmospheric Observatory, on the north Norfolk coast in the UK.

The estimate uses a new method for separating CO₂ signals from land plants and fossil fuels in the atmosphere.

Previously it has not been possible to quantify changes in regional-scale fossil fuel CO₂ emissions with high accuracy and in near real-time.

Existing atmospheric-based methods have largely been unsuccessful at separating fossil fuel CO₂ from large natural CO₂ variability, so that estimates of changes, such as those occurring in response to the lockdowns, must rely on indirect data sources, which can take months or years to compile.

The atmospheric O₂-based method, published in the journal Science Advances, is in good agreement with three lower frequency UK emissions estimates produced during the pandemic by the Department for Business, Energy and Industrial Strategy, the Global Carbon Budget and Carbon Monitor, which used different methods and combinations of data, for example those based on energy usage.

Crucially, as well as being completely independent of the other estimates, this approach can be calculated much more quickly.

The researchers are also able to detect changes in emissions with higher frequency, such as daily estimates, and can clearly see two periods of reductions associated with two UK lockdown periods, separated by a period of emissions recovery when Covid restrictions were eased, during the summer of 2020.

Researchers at UEA - home of the UK’s only high-precision atmospheric O₂ measurement laboratory - worked with colleagues at Wageningen University in the Netherlands and the Max Planck Institute for Biogeochemistry, Germany.

The study’s lead author, Dr Penelope Pickers, of UEA’s Centre for Ocean and Atmospheric Sciences, said: “If humans are to reduce our CO₂ emissions from fossil fuels and our impact on the climate, we first need to know how much emissions are changing.

“Our study is a major achievement in atmospheric science. Several others, based solely on CO₂ data, have been unsuccessful, owing to large emissions from land plants, which obscure fossil fuel CO₂ signals in the atmosphere.

“Using atmospheric O₂ combined with CO₂ to isolate fossil fuel CO₂ in the atmosphere has enabled us to detect and quantify these important signals using a ‘top-down’ approach for the first time. Our findings indicate that a network of continuous measurement sites has strong potential for providing this evaluation of fossil fuel CO₂ at regional levels.”

Currently, fossil fuel CO₂ emissions are officially reported with a ‘bottom-up’ approach, using accounting methods that combine emission factors with energy statistics to calculate emissions.

These are then compiled into national inventories of estimated greenhouse gas (GHG) emissions to the atmosphere from anthropogenic sources and activities, such as domestic buildings, vehicles, and industrial processes.

However, inventories can be inaccurate, especially in less developed countries, which makes it more difficult to meet climate targets.

It can also take years for the inventory assessments to be completed, and at the regional scale, or on a monthly or weekly basis, the uncertainties are much larger.

An alternative method of estimating GHG emissions is to use a ‘top-down’ approach, based on atmospheric measurements and modelling.

The UK emissions inventory is already successfully informed and supported by independent top-down assessments for some key GHGs, such as methane and nitrous oxide.

But for CO₂, the most important GHG for climate change, this has never before been feasible, because of the difficulties distinguishing between CO₂ emissions from fossil fuels and land plant sources in the atmosphere.

Dr Pickers said: “The time taken for inventories to be completed makes it hard to characterise changes in emissions that happen suddenly, such as the reductions associated with the Covid pandemic lockdowns.

“We need reliable fossil fuel CO₂ emissions estimates quickly and at finer scales, so that we can monitor and inform climate change policies to prevent reaching 2°C of global warming.

“Our O₂-based approach is cost-effective and provides high frequency information, with the potential to provide fossil fuel CO₂ estimates quickly and at finer spatial scales, such as for counties, states or cities.”

The team used 10 years of high-precision, hourly measurements of atmospheric O₂ and CO₂ from Weybourne Atmospheric Observatory, which are supported by the UK’s National Centre for Atmospheric Science. Having long-term measurements of these climatically important gases was crucial to the success of the study.

To detect a Covid signal, they had to first remove the effects of atmospheric transport on their O₂ and CO₂ datasets, using a machine learning model.

They trained the machine learning model on pre-pandemic data, to estimate the fossil fuel CO₂ they would have expected to observe at Weybourne if the pandemic had never occurred.

They then compared this estimate to the fossil fuel CO₂ that was actually observed during 2020-2021, which revealed the relative reduction in CO₂ emissions.