Aerial detection of insect damages in mixed-hardwood forests

To overcome current limitations associated with ground and satellite-based forest health surveys, this study was conducted to develop a field application package including unmanned aerial systems (UAS, a.k.a. drone), sensors, and field operational protocols that can guide forest managers and make forest health monitoring safer, more accurate in real-time, and more economical. The design concept of the UAS was small, versatile, low-cost, and easily transportable such that flight operations could be handled by a small crew with minimal logistical overhead and operational cost. We conducted a series of field tests for remote sensing in the Allegheny National Forest. We selected known mixed wood stands for various pest damage/crown dieback. A series of tests for sensor’s detectability was conducted. The UAS will be flown at five different altitudes (i.e., 50, 100, 200, 300, and 400 ft.), and remotely sensed data including aerial images were taken from the cameras and sensors. We investigated hyperspectral data, thermal data, and normalized difference vegetation index (NDVI) to determine the detectability of the sensors at different flight altitudes. The results of the tests showed that various gradients of crown dieback caused by different agents including emerald ash borer, peach bark beetle, cherry leaf spot, and other factors. This study established the relationship between optimal altitudes of UAS flight and sensor’s detectability for aerial survey of pest damage or crown dieback.