Forest Health Aerial Detection Survey Methodology
Each year, the U. S. Forest Service’s Forest Health Protection (FHP) program and its State Partners, the Oregon Department of Forestry (ODF) and the Washington Department of Natural Resources (WDNR) collaborate to monitor the health of all forested lands across the Pacific Northwest Region. The Pacific Northwest Region includes Oregon, Washington, and a small piece of northern California. Forest health conditions have gained popular attention in recent years because of environmental concerns associated with changes in long-term weather patterns, outbreaks of forest insects and diseases, population growth, and long-term resource management. If significant forest damage occurs from insect infestations, disease, human-made stresses, or abiotic causes such as wind or heat scorch, FHP is typically one of the first groups to identify where damage is, how severe it is, and help to determine the probable cause of the forest damage observed. Forest Health Protection then shares these observations of forest damage with the public and/or appropriate land managers for any given area. With its state partners, FHP works with all federal, state, and private landowners to assist with forest health concerns. Forest Health Protection is one of the only organizations that monitors the health of all forested lands regardless of ownership or administrative boundaries to get the whole picture of the forests’ current state.
There are approximately 47 million acres of forested land across Oregon and Washington. One may wonder, how does the U. S. Forest Service (USFS) monitor such a vast amount of forested terrain and diagnose all the different causes of forest damage each year? The USFS has used an evolving system of monitoring and diagnosing forest health conditions for over 100 years. It starts with the forest health specialists who are trained in the fields of entomology, plant pathology, ecology, forestry, and tree ecophysiology. Over time, these specialists have observed and recorded patterns of forest damage that affect different tree species. They closely monitor how different tree species are affected by injuries, noting subtle differences between the different causal factors. Specialists are able to differentiate what is happening across the forested landscape. This knowledge has been passed down, through training and demonstration or long-term measurements on the ground, creating a solid foundation on which forest health specialists continue to build. This knowledge is then applied to a variety of monitoring methods from site visits on the ground, staffed aerial surveys, to unmanned aviation systems (UAS) or drones, or by working with imagery collected by satellites. Most of the monitoring at the landscape scale is accomplished by staffed aerial detection surveys.
Staffed aerial detection surveys (ADS) provide an annual snapshot of forest health conditions over large areas. Most aerial surveys are conducted in fixed-wing aircraft, such as small planes, while some special surveys are conducted using helicopters. Aerial surveys are more efficient, reliable, timely, and economical than any other operational remote sensing methods to date. Drones or UAS, that are available to the USFS, can only cover a few thousand acres at a time and satellite imagery may come with a suite of challenges, such as availability, cost, resolution, clouds, mountain shadows, and timing/angle of imagery that currently hinder their applicability when compared to staffed aerial detection surveys that provide information on a quicker turnaround. In the US, forest health aerial surveys started in the 1920’s and have been consistently conducted in the Pacific Northwest since 1947. Each year, since those early days, the aerial detection survey improves upon the last by modifying aircraft, passing along knowledge, and developing customized tools to assist in the monitoring of America’s forests. Aerial surveys can be used as the first step with coarse estimates of recent damage across vast areas as part of a multi-tiered approach for detection, evaluation, and research. Then ground sampling and other remote sensing techniques can be utilized to gather additional data on changes in forest conditions. Data collected during these surveys has proven useful in early detection of invasive species, rapid response actions such as quarantine, control, or eradication. The effects of endemic pest outbreaks are also recorded and provided to national and local area land managers, referenced in a variety of conditions and congressional reports to aid decision making.
To conduct the staffed aerial detection survey (ADS), forest health specialists board small aircraft equipped with a digital aerial sketch-mapping system that incorporates tablets, geographic information system (GIS), and global positioning system (GPS) technology into a single software application called Digital Mobile Sketch Mapper (DMSM). The forest health crew flies over the forests at 500 to 3000 feet above the forest canopy in either a grid or along ridges in a contour pattern depending on the terrain below. Traveling around 100 mph and looking out about two miles, each observer can survey an estimated 15-30 acres per second, diagnosing forest damage by tree species, mapping its extent, and rating the severity of that damage costing less than a penny per acre to survey. Forest health specialists can distinguish the subtle differences between tree species and the subtle differences between the variety of possible damage causing agents which could be impacting each tree species.
For example, let’s imagine a large ponderosa pine forest. As we fly over this forest, specialists are looking out the windows for different forest characteristics such as color and condition of the canopy to determine if the trees are healthy, defoliated, dying, or recently dead depending on the appearance of tree crowns. As a ponderosa pine dies, typically the foliage will transition from a characteristic green to a specific chlorotic yellow - straw color, then over time change to an orange-reddish color. The distribution of the dying or recently dead ponderosa pines can tell us more about the possible cause of this forest damage. If the recently dead pines are all the same color of red and those dead trees only extend about 100 feet from the highway, but the rest of the forest is healthy and green, then deicing salt or herbicide damage from road management operations is most likely the cause. Now imagine that we fly further into the wilderness away from roads and observe clusters of dead ponderosa pines in a variety of fading canopy colors scattered throughout the forest. Bark beetles or certain diseases may be causing patchy tree mortality.
The same process is used for detecting defoliation events in ponderosa pines and, based on the topographic position of the trees and which needles (interior older foliage only, exterior youngest foliage only, top-down, bottom-up) are affected, specialists can make an educated guess from the plane whether it is most likely caused by pandora moths, pine butterflies, sawflies, or fungal pathogens that are damaging those ponderosa pine needles. This process of distinguishing the subtle differences repeats for each tree species and location as we fly along recording observations on the DMSM app on tablets which is keeping track of the plane’s location while moving on a customizable, scrolling map. Using a variety of aerial imagery as a background for this map, surveyors can correlate the damage they are observing out the window with where the damage would be located on the digital map displayed on the tablet. Typically, aerial survey data are collected and processed from the summer to the fall to capture the most forest damage signatures as possible, then ideally the flights are completed by October 1st.
Since forest pests and the damage they cause are dynamic and highly variable each year, the resulting data will also be highly variable. No two ADS observers can be expected to record the same forest damage exactly the same way. For this reason, aerial detection survey data should be regarded more of an interpretation than an exact representation. It is important at the outset that this be understood, not only by the conscientious aerial observers who find that their data may not be in close agreement with their peers or with a precise aerial photo survey, but also by the forest manager, who may want to put the information to use. Staffed aerial surveys are highly subjective, and the resulting data depend on the competence of the aerial mapper and the conditions under which the data were obtained. Because of these circumstances, staffed aerial detection survey data should be regarded as a coarse "snapshot" of landscape level forest health and/or forest change condition. Spatially, the data are best displayed at relatively fine scales such as 1:100,000, 1:250,000 or 1:500,000. The data are better used for demonstrating coarse estimates or trends rather than exacting precise measurements.