The Northeastern Forest Health Atlas, or NEFHA, is a compilation of state and federally-collected aerial detection surveys (ADS, recently renamed IDS for insect and disease surveys) for the Northeastern US. The NEFHA interface provides maps of forest damage events overall and by specific damage types and causal agents. NEFHA users can filter by causal agent, damage type, state, and year, as well as view graphs and tables of disturbance patterns over time, and download maps and data. Although other online forest damage databases exist, they lack the ability to explore disturbance events over variable time spans, do not include all causal agents, have fewer years of data available (which precludes the examination of long-term patterns), and/or display disturbances by county rather than by mapped polygon extent.

The FEMC will continue to add annual aerial survey data as they become available, as well as digitize historical maps to extend the available record. All of the states in this region have forest damage maps dating back to the mid-20th century that are being prioritized for digitization and quality control, thus providing a longer-term record of a variety of causal agents for future assessments. NEFHA also includes locations of field-based forest health related research projects with links to project information and datasets for future investigations linking aerial survey to field data.

Aerial sketch-mapping has been conducted via fixed-wing aircraft by state forest health programs on state and private land, and by the federal agencies on federal lands for many decades. These aerial surveys are aligned with the US Forest Service Forest Health Monitoring (FHM) Program and reporting standards and supported by the USFS Forest Health Assessment and Applied Sciences Team (FHAAST – formerly Forest Health Technology Enterprise Team).

During aerial surveys, forest health technicians digitally delineate areas of visible forest damage. Resulting damage polygons include the following attributes: damage type, causal agent, year of occurrence, and, in some cases the severity of forest damage, dominant forest type affected, and host species for pests and pathogens. Under this schema, damage. agents can be classified as one of 11 disturbance types. Areas with unknown or uncertain cause of damage are ground-truthed for accuracy. However, if it is not possible to confirm agent, the causal agent is entered as unknown.

Working with personnel from the five state forest health agencies, we combined the aerial survey datasets for the region in ArcGIS (version 10.4). We standardized the aerial survey data from each state according to the Aerial Survey Geographic Information System Handbook (USDA Forest Service, 2005). We aggregated the state data into a regional geodatabase, conducted quality control checks, and backfilled missing records with annual federal files from the USFS Northeastern Area State and Private Forestry Group (NA). We standardized damage-causing agent codes by reclassifying those codes that did not correspond to those provided by FHAAST (i.e., values <-1 or between 0 and 10000) using existing NA datasets and notes in the dataset. For rows that did not correspond with NA data, we replaced codes that did not align with FHAAST with the code 99999 for ‘unknown’. We dissolved aerial survey polygons by year in order to avoid overlapping polygons mapped by adjacent states.

While aerial survey data provide long-term, spatially explicit damage observations, they are not comprehensive of all forest damage that has occurred. There are a number of important caveats of these data that must be understood before using them:

1. The type of damage affects how easily it can be observed from the air. Subtle damage agents or those that have differing effects on trees within a mixed forest stand, like diseases, minor wind storms, or light decline symptoms, may be overlooked during aerial surveys. For example, severe defoliation and recent mortality are easier to delineate in aerial surveys than minor branch breakage.
2. Aerial assessments are typically conducted only once during the growing season, although some state agencies may respond opportunistically to known damage events at other times. If damage occurs too long after the aerial assessment or recovery is quick, the damage may be overlooked.
3. States have responded to forest damages with differing priorities and goals. Causal agents that affect timber production and/or result in large-scale, conspicuous damage, such as that caused by insect defoliators and sap feeders like balsam woolly adelgid and spruce budworm, often have been prioritized in aerial surveys.
4. Aerial sketch-mapping flight paths have not always covered entire states, instead varying with agency priorities and budgets.
5. Damage agents that have small polygonal delineations (<1,000 ha) tend to have been identified less frequently. While this may provide information on the nature of small-scale disturbance events, it may also suggest that small damage polygons are more difficult to identify or were not prioritized in aerial surveys.
6. Damage area attributed to some agents may be overestimated; it is often easier to ‘lump’ areas of damage together while sketch-mapping.
7. Severity of damage is often not uniform throughout a polygon; polygons can include many unaffected trees while simultaneously underestimating the per unit area severity of some agents. Because damage extent is estimated from the air, polygon edges also can erroneously extend into water or other non-forested land.

NEFHA can be used to examine how forest damage has changed over time, quantify the approximate amount of damage from an agent, agents, or by damage type, and identify pertinent research. It can be used for education, or to assist with resource management and planning. Please make sure you have read the important caveats of these data.