Forest Investment Account (FIA) - Forest Science Program
FIA Project Y092213

    Integrating Historical Air Photo Data for Cedar Analyses
Project lead: Forest, Marguerite (Haida Mapping)
Contributing Authors: Forest, Marguerite S.E.; Reynolds, Nick; Wilhelm, Lana; Fortier, Jean-Marc; Kregosky, Roy
Subject: Forest Investment Account (FIA), British Columbia
Series: Forest Investment Account (FIA) - Forest Science Program
The main forest management object of this project is growth and regeneration of sufficient red and yellow cedar of monumental quality for Haida cultural uses. First Nations in coastal British Columbia make extensive use of large, high-quality cedars for long-houses, canoes, totem poles, planks, and other objects. On Haida Gwaii, evidence suggests that there are now very few red and yellow cedar in second growth stands, that cedar regeneration is impeded in part due to severe over-browsing by introduced deer, and that remaining cedars are not of as good quality.

Large high quality cedars are now a key focus for First Nations consultation and accommodation on Haida Gwaii. They are also a critical indicator in the co-managed Land Use Plan. This research will help determine changes in the distribution, abundance, and quality of red and yellow cedar on Haida Gwaii between 1937 and 2007, as well as the best sites for future growth of high quality cedar. Project results will be used to develop a registry of monumental cedar for Haida cultural use and to establish a series of age graded sites for long-term regeneration of high quality cedar. The registry and regeneration sites will bring a higher level of certainty to local timber harvests.

Time series data are the key to more precise understanding of long-term forest cover changes and resulting variation in timber growth and value. Sources of time series data include historical records, field surveys, and remote sensed data. Haida Gwaii is fortunate to have all three. Despite the availability of these data, long-term changes such as shift in species composition, reduction in tree sizes, and deterioration in wood quality have not been rigorously quantified for cedar on the islands.

Some of the first air photos in coastal British Columbia were flown for Haida Gwaii in 1937 at ~1:15,000. The photos are of good quality and provide detailed information. They are now being scanned, orthorectified, and mosaiced in a series of projects funded by the South Moresby Forest Replacement Account (SMFRA), Gwaii Trust, and other sources. A trial has been done on using these photos to interpret forest cover and pinpoint locations of individual large trees on Lyell Island. No other sites have been examined so far. The method for that project was desktop stereo pair interpretation. Forest cover polygons and attributes were delineated by hand on mylar overlays. Mylar overlaid photos were scanned, georeferenced, and orthorectified, then the polygons were digitized on screen and their attribute data entered into GIS files. Results were checked against 1960s forest cover maps and cruise data, but not against any field data. It was a time-consuming procedure, involving much duplication of effort, and without any of the field samples now required for accuracy assessment.

New forest cover interpretation methods focus on 3D digital stereo viewing using specialized software and equipment, such as the PurView extension for ArcGIS. On-screen digitizing is done from scanned photos. The need to scan, georeference, and orthorectify both original and mylar overlaid photos, and to delineate polygons both on mylar overlays and on screen, is eliminated. Use of these techniques would speed up historical forest cover interpretation significantly. The first goal of this project is to test if these new methods can be used effectively on the 1937 air photos. In 2004, Vegetation Resource Inventory (VRI) mapping standards were updated to reflect use of digital “softcopy” technology and the RISC format. The new standards will be followed as much as is practicable for the 1937 air photos, and Haida Mapping technicians will be trained in these interpretation methods.

Timber cruise data and forest cover maps for Haida Gwaii date back to the 1960s. Subjective back-casting is needed to compare these data to interpretations from the 1937 air photos. So far, no other method has been used to check the Lyell Island sample 1937 forest cover. Interpretation of the same site by a different interpreter is no longer adequate, and a Ground Sampling Phase is now required for verifying VRI mapping. The second goal of this project is to check 1937 air photo interpretations against field data. Data from existing field sample sites will be emphasized, but additional field surveys will be required to complement these data. New surveys will focus on locations needed to improve the spatial distribution of field sites and on new data needed bring existing field data sets up to VRI standards.

Biogeoclimatic Ecosystem Classification (BEC) sample sites across the range of variants and site series relevant for red and yellow cedar will be the primary field data used for this test. Other field data sets will be used where no BEC sites are avalable. Where the selected field sites were logged, field surveys of stumps will be conducted to determine the species, ages, basal areas, etc. of trees that grew there in 1937. Haida Heritage and Forest Guardians (HHFG) field crews have extensive experience in surveying for Culturally Modifed Trees (CMTs) and monumental cedars. They will be trained in the additional Ministry of Forests and Range (MoFR) Revenue Branch cruising techniques needed for VRI and stump surveys, including dendrochonology methods. Where possible, the field data sets will be compared to the earliest timber cruise data and forest cover maps from the 1960s to confirm forest cover at the middle of the 1937-2007 time series.

The third goal of this project is to analyze changes in cedar characteristics at different BEC sites by comparison with modern imagery. New color air photos of Haida Gwaii from 2007 at 1:20,000 will provide the best match to 1937 air photos. Haida Mapping GIS/RS technicians will be trained in these modern interpretation methods also, and forest cover derived from this imagery will be checked using field data from the same BEC sites
Related projects:  FSP_Y081213

Forestry Extension Note (0.4Mb)

Updated August 16, 2010 

Search for other  FIA reports or other Ministry of Forests and Range publications.

Please direct questions or comments regarding publications to