Blog

Why are new estimates of emissions from tropical deforestation lower?

  • 26 Jun 2012, 14:25
  • Freya Roberts

A new study has suggested that emissions from tropical deforestation could be lower than previously thought. So what's changed? Were older estimates wrong, and if so, what might this mean for the future of the planet's atmosphere?

New measurements from satellites

The new estimate comes from improved satellite measurements of how tropical forest cover is changing. This ability to estimate forest loss from orbit is relatively new, and allows scientists to calculate emissions from deforestation with greater certainty.

By comparing older and newer satellites images showing forest cover, scientists are able to work out two things: forest cover and the amount of biomass in an area. Assessing the amount of ground covered by trees alongside the amount of biomass in an area tells you about the amount of carbon stored in that area - the 'carbon stock', and this lets researchers roughly work out the amount of carbon released through deforestation.

This study, and others in recent years, calculate gross emissions - making no assumptions about what the land is used for afterwards and how this affects overall changes in emissions from deforestation. This removes uncertainty in the calculation, but it makes it tricky to compare new estimates to older studies, that did look at land use change.

Older studies had different uncertainties

Instead of using satellite measurements, older estimates of emissions from deforestation relied upon countries to report the amount of deforestation that had occurred to the United Nations Food and Agricultural Organisation (FAO), which collated the data. Using this data and a computer model, researchers calculated the emissions released from deforestation.

In practice, self-reporting often didn't work that well - estimates were often inaccurate and the data was collected inconsistently by different countries.

Some of the assumptions made to simplify calculation of emissions also introduced broad margins of error. For example, models tried to work out what happened to land after it was deforested, and feed that into the carbon accounting, but there often isn't much data to show what land is used for after deforestation, increasing uncertainty in the estimates.

Satellite measurements help address some of these problems, and avoiding looking at land use after deforestation may be a better approach to a situation where there often isn't good data.

The new results in the context of other studies

Compared to other studies, estimates of carbon emissions from this new study are considerably lower, and it's not entirely clear why.

As seen from the table below, the estimate calculated by the authors is around a quarter of the Pan et al. estimate, and around a third of the Baccini et al. estimate.

Both of these earlier studies also use satellite images, but combined them with FAO data. It may be that this introduced inaccuracies into the calculations.

Source Gross Carbon Emissions
(petagrams of carbon per year)
Time period
Pan et al., 2011 2.8 2000 - 2007
Baccini et al., 2012 2.2 2000 - 2010
Harris et al., 2012
(This study)
0.81 2000 - 2005


Estimates of deforestation emissions from studies using satellite data. Source: reproduced from Harris et al., 2012 (Supplementary Information Table S1). 1 petagram = 1,000,000,000 tonnes.

We contacted the lead author, Dr. Nancy Harris, to ask what could explain the difference between the estimates.

She suggested it was down to the contrasting methods used to calculate emissions. Using satellites to track emissions, she argued, is a "more consistent methodological approach":

"Satellite data allowed us to pinpoint more precisely where deforestation was occurring and what the resulting emissions were."

Even so, the magnitude of the difference is still quite surprising, and suggests more research to figure out what it is that's causing the estimate to be so different.

Having accurate estimates of deforestation and the emissions it causes is important, not least because there are a number of climate policies structured around an assumption that we can measure deforestation accurately. If research concludes that emissions from deforestation aren't as high as we thought, it might mean a rethink.

Email Share to Facebook Stumble It
blog comments powered by Disqus