Seeing The Forest Through The Trees

Sampling secondary structure in pine stands with novel aerial photography

How will beetle-attacked pine stands look in the future?

Beetle-attacked lodgepole pine stands in the B.C. Interior are being harvested in order to try to utilize the wood before it loses its economic value. These forests will be managed for rapid reforestation. However, the epidemic is so extensive that thousands of hectares of attacked pine stands will remain on the landscape. What will these forests look like in the future?

How a beetle-attacked pine stand might change over time
The way that these stands develop over time is the result of the combined effect of deterioration of dead trees and the growth of new trees. Deterioration rates depend on factors such as snow loading, and the chance occurrences of windstorms and fire. Growth depends on what’s there already and the locations of seed sources. Many pine stands are even-aged because they regenerated after a stand-destroying fire. However, other stands have more complex histories which have resulted in gaps, uneven tree ages, and a mix of tree species. Also, this beetle epidemic has not killed every pine tree. Mortality has been very high in many stands but other pine stands will have varying numbers and sizes of conifers that survive this epidemic.
When the overstory dies, the physical environment on the forest floor changes, thereby affecting the growth of understory trees. Changes include more water and light being available. Light levels in deteriorating and growing pine stands are being documented as part of this project. Effects of canopy mortality on snow accumulation and ablation

Secondary structure is the future forest

Conifers that survive the beetle epidemic will tend to be the smaller pine trees and the non-pine species. However, some mature lodgepole pine will also survive. These are collectively termed “secondary structure”, (Coates et al., 2006). As those authors noted, the beetle epidemic has increased the value of secondary structure in pine stands because when the pine overstory dies, any remaining live trees will benefit wildlife habitat, accelerate hydrologic recovery, and could provide mid-term timber supply. Unfortunately, there is a dearth of inventory information on understory conifers, partly because understory trees are almost impossible to see on regular aerial photographs. Forest inventory information that cannot be mapped on standard aerial photos or satellite imagery is often obtained by sampling. The most accurate sampling method is ground-based surveys but this is slow and expensive. This project is investigating a new sampling method that would operate at a scale between ground-based and standard remote methods.
This project is being supported by
Natural Resources Canada’s Mountain Pine Beetle Program and by the B.C. Ministry of Environment’s Ecological Restoration  program

The Novel Photography

Vertical aerial photos taken when snow covers the forest floor and when the sky is overcast provide a unique view of forest structure. This combination of lighting conditions provides a diffusely backlit silhouette of all vegetation and woody debris that is not covered by snow. Aerial photographs of the same research plots taken under different snow and lighting conditions show the clarity with which this method can reveal both overstory and understory trees.

Complete snow cover
and clear sky.
Complete snow cover
and overcast sky.
Partial snow cover
and overcast sky.

Here are aerial photos of a 50 by 50 metre plot in a 100% pine pine stand. According to the provincial inventory, this stand is 135 years old, 22 metres tall, and has 650 stems/ha (B.C. Vegetation Resource Inventory). These images show that lighting has a radical effect on the visibility and appearance of trees. Under clear skies, large scale aerial photos of trees are dominated by patterns of light and dark due to uneven illumination (first photo). This is not a problem if one is interested  mainly in the canopy-forming trees. Indeed, traditional aerial photography normally requires cloud-free conditions due to flight altitudes. When the sky is overcast, the distraction of shadows is removed but these photos reveal another large difference. The appearance of the trees is very different depending on whether they are seen as silhouettes against a white background (second image), or illuminated from the front against a dark background (third image). On the image taken when there was complete snow cover and overcast skies, understory spruce trees are clearly visible under the pine overstory. This is enhanced by the fact that understory spruce tend to have crowns that are much larger in diameter and denser than those of pine trees – especially dead pine trees, and by the fact that pine stands in the Interior tend to have relatively low crown closures.

The effects of different lighting conditions on the ability to detect different scene features are well-known in the fields of microscopy and machine vision. In contrast, unusual scene illumination has not been well-utilized in photographic remote sensing of forests because the lighting cannot be controlled. Aerial photography under these conditions must be done opportunistically.


The research project

Aerial photos taken in late winter and early spring of 2006 for another project ( Effects of canopy mortality on snow accumulation and ablation  ) provided the initial sample of images. In March and April 2007, new aerial photos were acquired with funding from the Ministry of Environment’s Ecological Restoration Program. In early 2008, VRI Interpreters were contracted to create stem maps and measure crown diameters of live understory conifers on a subset of these photos in order to test the usability of the imagery, the new interpretation process, and two stem mapping platforms (Ozi Explorer and Google Earth).

Example of georeferenced aerial photo and stem map of live understory conifers.    
2007 aerial photo overlaid in Google Earth One-quarter hectare plot superimposed on aerial photo. Stem map of live understory based on photo interpretation.

This Google Earth image overlay can be downloaded here 695 full res.kmz  

Project participants include the University of Victoria, West Fraser Mills Ltd., Williams Lake Woodlands, and B.C. Timber Sales, Williams Lake.

In 2007, new research plots were selected where secondary structure will be sampled by two different methods for comparison. Aerial photos will be acquired using the new method and will be used to map and measure secondary structure as in the above example. Secondary structure will be mapped and measured in the same plots in the field and the two maps and sets of measurements will be compared.  



This MoFR website has links to documents pertaining to stewardship in the context of the Mountain Pine Beetle epidemic, including guidance from the Chief Forester and information on secondary structure.


Coates, K. David, Craig DeLong, Philip J. Burton, and Donald L. Sachs. 2006. Abundance of Secondary Structure in Lodgepole Pine Stands Affected by the Mountain Pine Beetle. Report prepared for the Chief Forester of British Columbia. Available at MoFR website

For More Information

Contact: Pat Teti
Research Hydrologist
Southern Interior Region
200 – 640 Borland Street
Williams Lake, B.C., V2G 4T1
Phone: 250 398-4752, eMail: