Monitoring our forests can be difficult from the ground, but satellite technology offers a bird's eye view, writes Danielle Barron.
If you can't see the forest for the trees, then change your perspective. Try spotting the forest from a satellite.
Researchers in University College Dublin's school of biology and environmental science are taking this approach, working to provide reliable, up-to-date statistics on the extent and composition of forest resources using satellite images.
The Irish forest estate is constantly changing due to planting and harvesting, so better monitoring techniques are required, says Daniel McInerney, a PhD candidate working under Prof Maarten Nieuwenhuis at UCD.
"Over the past two decades, there has been an increase in private afforestation, but still very little information is known about this component of the estate," he explains.
The problem of monitoring forests can only get worse, if the ambitious national target of 17 per cent forest cover by 2035, up from about eight per cent today, is achieved.
The research project to overcome the monitoring problems began in 2006 after a successful application to the Irish Research Council for Science, Engineering and Technology (Ircset) under the Embark initiative. The Irish Forest Service also agreed to co-operate.
"The National Forest Inventory (NFI) was underway, so it was a good time to begin researching methods that could complement and enhance this dataset," recalls McInerney.
As the NFI was designed for national reporting, it did not specifically address the regional or local scene. Also, field-based inventories are expensive and time-consuming, he says.
Yet accurate monitoring tools for production planning and forecasting, biodiversity assessment, identifying sources of renewable energy, amenity provision, landscape planning and assessing the forest's role in the carbon cycle are essential, McInerney maintains.
He decided to take a different approach using remote sensing, something he came across while abroad. "During my studies I was very fortunate to spend some time at the Finnish Forest Research Institute, where I was exposed to new forest inventory methods," he explains.
He chose to modify a technique with the most unlikely title, "k Nearest Neighbour" (kNN), a statistical pattern recognition method. While it has been used for some time by Finland and Sweden, there had been little research applying it to the temperate forest conditions in Ireland and Britain, says McInerney.
KNN involves combining field inventory data from the NFI, satellite imagery and other datasets such as landcover and topography data. It also requires a "statistical estimator" to produce a digital thematic map showing, for example, forest type and volume per hectare.
McInerney's three-year research project uses kNN on two pilot regions in counties Clare and Wicklow, for which complete field datasets exist. Satellite imagery from the Landsat 7 satellite and more recent imagery from the French SPOT satellite have been used to produce an up-to-date forest mask.
Refinements of this mask will produce digital thematic maps that identify the spatial distribution of forest resources, he explains. These maps will show the distribution, say, of conifer versus broadleaf trees. "Spatially referenced information in the form of digital thematic maps is extremely useful as they can subsequently be integrated into a Geographic Information System (GIS) and used for further spatial analysis and mapping," says McInerney. "Spatial analysis can be used, for instance, to identify all forests of a certain age or species within a specified distance or travel time of a sawmill."
The fact that fresh satellite imagery is being acquired for the entire country at regular intervals also means that quantitative records of the forest estate can be kept up-to-date in the future.