A small patch of wetland on the shore of Maryland's Chesapeake Bay could hold the key to some of the most pressing concerns about the effects of global warming.
Every year the world loses thousands of hectares of land as rising sea levels engulf coastal areas.
By the end of this century the sea is expected to rise another 18-59cm (7-23in), threatening heavily populated coastlines such as Bangladesh and entirely destroying other nations such as the Maldives, in the Indian Ocean.
Most of these projections are based on theory as nobody knows exactly how the earth will react to climate change.
But in the world's longest running carbon dioxide experiment in Maryland, scientists have created a unique salt marsh that simulates the environment of the future.
And they've discovered that while increased carbon dioxide (CO2) is the biggest cause of global warming, it also makes some plants grow faster which builds new soil and helps the land keep pace with rising sea levels.
"That's the silver lining," says Patrick Megonigal, a senior biogeochemist at the Smithsonian Environmental Research Center. "CO2 acts as a fertiliser on some plants, and in a marsh like this, a faster growing plant has some good characteristics.
"This marsh can actually build soil through root growth and more soil means this marsh can rise upwards and therefore keep pace with rising sea levels."
The marsh is dotted with atmospherically controlled chambers that contain the same amount of CO2 that the planet may be exposed to by the year 2100 - roughly double what it is today.
"They're like time capsules. We are simulating the future inside them," says Dr Megonigal. "We're trying to travel forward in time by subjecting these plants to the conditions the whole world will be subjected to a hundred years from now."
But there are other man-made factors that may counteract soil creation. Pollution - particularly from nitrogen - can have a major impact.
"We found that when you combine nitrogen with CO2, the positive effect on soil elevation is cut in half. So while increased CO2 helps the marsh keep pace with rising sea levels, increased nitrogen seems to work in the other direction. What you give with one hand you take with the other. It's a very complex situation."
Coastal wetlands are the first defence against climate change and the 60-hectare (148-acre) salt marsh at the heart of the Smithsonian Environmental Research Center has been home to some of the most important ecological studies of the past 40 years.
Scientists here were among the first to take seriously the effects of global warming and begin studying its impact.
In another experiment at the centre, one part of the marsh has succumbed to Phragmites australis, a plant better known as the common reed.
Although it is native to North America, another strain was introduced from Europe in the 1800s and is classed as an invasive species. It can grow to 20ft (6.1m) high and form an impenetrable wall of bamboo-like stalks.
"My research addresses how global climate change will affect the invasion process in natural eco-systems," says environmental scientist Thomas Mozdzer. "Preliminary experiments have shown that increases in CO2 and nitrogen could triple the plant's growth."
Such growth can critically harm native habitats, displacing native plants and the organisms and animals that need them to survive. By understanding the effects of rising CO2, scientists may be able to discover how to control species such as Phragmites australis.
"Knowing how these eco-systems behave in the future is useful for planning," says Dr Megonigal. "Certain invasive species may degrade the marshland in terms of habitat, but their accelerated growth may help offset the effect of rising sea levels. If we better understand that trade-off we can decide whether to allow the plant to invade."
The ongoing carbon dioxode experiment began in 1987 when scientists started examining whether plants were able to extract and store CO2 as levels in the atmosphere continue to rise.
"There's lots of interest in creating marshland to sequester CO2. The type of work carried out here will find out whether carbon stored today will still be there in a hundred years from now," says Dr Megonigal.
The experiments at the Smithsonian's Environmental Research Center are unique because they've been conducted continuously over such a long period of time. The marsh also has an electricity supply and platforms that make it accessible to scientists and easy to install equipment.
The latest results from this living laboratory were published internationally in the summer and the National Science Foundation has agreed to fund research for another decade.
"We've been doing this work for a quarter of a century," says Dr Megonigal. "That kind of long term data gives us something that short term experiments can't.
"An experiment that lasts two or three years tells us very little. The long term responses are what we need to project beyond 10 years, to say 100 years, which is our goal here."