NASA Terra Aqua MODIS: UK scientists reveal Amazon rainforest satellite imaging problem - Video

US Scientists have been puzzled by the apparent ‘greening up’ of the Amazon forest during its annual dry season.

However, Swansea University scientists, in research carried out in conjunction with NASA, have found that the Amazon is not as green in the dry season as researchers previously thought, because a trick of the light skews the satellite images.

The research is published in the latest issue of the scientific journal Nature; doi:10.1038/nature13006

This natural-colour satellite image shows the importance of correcting for sun-sensor geometry. 

On the left side, sunlight is backscattered by the Amazon rainforest, creating the appearance of brighter green leaves in some areas. 

To the right, sunglint makes the dark waters of the Amazon River and surrounding flooded wetlands appear silver or white compared to the darker forest. 

 Credit: NASA's Earth Observatory

The finding of the research yeam will help scientists develop a more accurate picture of changes in the Amazon, which is more important now than ever, given the enormous role the Amazon plays in regulating carbon dioxide, and influencing climate change.

NASA's Terra and Aqua satellites make daily observations over the huge expanse of Amazon forests.

The research team were investigating why previous satellite images seemed to show that the forests became greener during the dry season each year from June to October.

More greenery indicates productive, thriving vegetation, which would not be expected at a time of limited rainfall.

The new research shows that:

• The apparent greening of the Amazon in the dry season is an illusion. The forest does not become greener at all.
• It just looks that way due to a combination of shadowing within the forest canopy and the way that satellite sensors observe the Amazon during the dry season, which can create false “hot spots” in images.

Dr Jackie Rosette

‌Swansea researchers, working with NASA, developed a mathematical model which predicts how a forest will be seen from space, and how leaf area can be measured.

One of the Swansea team, Dr Jackie Rosette, spent 2 years at Goddard Space Flight Center, working closely with NASA colleagues.

Blue colours represent areas in Amazon forests where sensors and models can overestimate the green-up of vegetation; white represents areas that lack forest cover. 

The map is based on a model that extends the sun-sensor correction to all pixels in the southern Amazon. 

Credit: Doug Morton and NASA's Earth Observatory

Professor Peter North from the Department of Geography at Swansea University, one of the authors of the research paper, said:

"The Amazon is so vast that it’s only from space that we can properly observe it, so it’s very important that the satellite data gives us as accurate a picture as possible.

Our model has helped to identify the flaws in the previous interpretation. As a result, we can be much more confident that what we are seeing is a real pattern across the Amazon, not a trick of the light.

Having an accurate picture of the Amazon is essential if we are to understand its key role in shaping the climate."

Aboard the NASA satellites are sensors called MODIS (Moderate Resolution Imaging Spectroradiometer) which measure the amount of infrared light reflected from the Amazon.

Scientists use the ratio of red and near-infrared light as a measure of vegetation.

Doug Morton, NASA’s Goddard Space Flight Center said.

"We think we have uncovered the mechanism for the appearance of seasonal greening of Amazon forests – shadowing within the canopy that changes the amount of near-infrared light observed by MODIS"

The research implies that the previous hypothesis of increased productivity during dry seasons is likely to be false, and Amazon productivity may be more limited by water availability than sunlight.

This is critical for predicting the response of the Amazon to future climate change.