Optimizing drone mapping of forest and fuel structure
Drone mapping of forest structure and fuels has the potential to expand the detail and accuracy of data used in management decisions, while also improving field crew safety and data collection efficiency. This study identified the optimal software settings for image processing that balance data accuracy and processing efficiency for mapping individual trees. Researchers found that optimizing these settings can reduce computer processing time and data storage needs by approximately 75 percent, without a noticeable loss in data accuracy. At these settings, the mapped trees produced stand-level estimates of basal area that were within the Forest Service’s allowable error for designing forest inventories. These findings point to the potential for major time and cost savings in data generation, storage, and processing time when operationalizing drone-based forest monitoring strategies.
This work builds on earlier efforts that have optimized the parameters for drone data collection while in the field.
“We are at the point where drones can fly roughly 100 acres per hour in the field and then the processing algorithms can extract all of the tree parameters in about six hours of computer time,” said Wade Tinkham, a research forester with the USDA Forest Service’s Rocky Mountain Research Station.
Using drone-based forest monitoring strategies represents a substantial leap forward by providing complete coverage of tree observations like the location, height, diameter, and crown base height, enabling spatially informed timber and fire management decision-making.
Vision Aerial Switchblade Elite drone system during a mapping mission on the Arapaho and Roosevelt National Forest in Colorado.