New methods for remotely detecting isolated wetlands

Isolated wetlands, which typically are small, seasonally inundated pools, are an important landscape feature in many parts of North America that provide habitat for plants and animals, and perform other ecosystem functions such as groundwater recharge. Managers who wish to consider isolated wetlands in their sustainable forestry practices encounter many inherent difficulties when attempting to find and map them because they often are obscure and methods to detect them at the spatial scales necessary for conservation planning and forest management have been time consuming, cost-ineffective, or too coarse-filter.

Using data from a low-relief managed forest landscape in the coastal plain of North Carolina, investigators with Clemson University, in association with Weyerhaeuser Company and NCASI, recently developed new remote sensing methods to aid in detection, management, and conservation of isolated wetlands. The investigators used high-resolution LiDAR elevation data to create custom local-relief models that emphasize curvature. Ground validation revealed that the models correctly classified isolated wetlands at 97 of 114 sites, a mapping accuracy of 85.1%. Commission and omission errors were 14.9% and 5.3%, respectively. The new methods are described in the following paper:

Leonard, P. B., R. F. Baldwin, J. A. Homyack, and T. B. Wigley. 2012. Remote detection of small wetlands in the Atlantic coastal plain of North America: Local relief models, ground validation, and high-throughput computing. Forest Ecology and Management 284:107-115.

Contact information