Hemispherical Photography in Support of Forest Inventory and Silviculture

Abstract

Hemispherical photography (HP) can be used in support of silviculture and forest inventory. At the tree level, HP can be used to estimate attributes such as tree architecture, crown dimensions, and volume of standing trees. At the stand level, there are relationships between indices extracted from HP and stand attributes such as density, basal area , canopy cover, and gap size. As an aid to silvicultural activities, the technique can provide valuable information for monitoring the effects of thinning and patch cuttings on regeneration processes and stand development. The HP technique is versatile, scalable, and the results compare well with other methods that are used to estimate canopy cover and light transmittance in forests. This chapter provides examples of applications of HP at both tree and forest stand levels. The strengths and weaknesses of the technique will be examined, and a comparison will be made with other techniques available in support of silviculture and forest inventory.

Appendix: Instruments not Based on the Fisheye Approach

In the main part of this chapter, HP and other optical ground-based methods using the fisheye approach were presented as potential techniques in support of forest inventory and silviculture for estimating canopy cover and light transmittance in forests. As has been shown, the use of HP for forest inventory and silvicultural applications is relatively new, but promising results can be obtained if the proper equipment is used and the technique is applied correctly. Before the advent of HP, however, several other techniques and instruments were available for estimating canopy cover and light transmittance in forests. The goal of this Appendix is to provide the reader with an overview of the main techniques and instruments that are not based on the fisheye approach and which may represent a valid alternative, if less detailed information is required about the structure and light regime of a forest canopy.

One of the most popular of these devices is the convex spherical densiometer, which was initially described by Lemmon (1956). This pocket-sized instrument measures canopy cover. It is designed for rugged field use and is comprised of a convex mirror divided into a cross-shaped grid of 24 squares. It must be held out in front of the body, about waist level, so that a clear view is given of the forest canopy without the observer shading the grid. Canopy cover is then estimated by counting the number of squares filled by vegetation. Baudry et al. (2014) found that densiometer readings were closely related to the relative light intensity assessed in overcast conditions, and suggested that the use of this inexpensive tool should be expanded. Conversely, some researchers have reported inconsistent results while using the convex spherical densiometer (e.g., Vales and Bunnell 1988; Cook et al. 1995). Despite these criticisms, convex spherical densiometers have remained a tool of choice because other cost-effective measurement devices are lacking (Forest Science Project 2000). However, according to Englund et al. (2000), spherical densiometers can be successfully used to characterize forest light environments as long as users have adequate experience and are trained to a consistent standard.

The angular densiometer is another instrument that is used to measure canopy cover. It consists of a hinged clamshell case containing a convex mirror, a bubble level, a magnetic compass, and a fixed eyesight. It is similar in design to a convex spherical densiometer but differs in that it provides a delineation of the sun path, a more accurate orientation relative to the celestial hemisphere, and a clearer image of the canopy (Teti 2001). It measures the percentage of time that shade is available during a 4-h period between 10:00 and 14:00, which is a period of time that have been judged to be critical for regulating stream temperature (Teti and Pike 2005).

The vertical tube is another device that is based on a similar principle for measuring canopy cover (Johansson 1985; Korhonen and Heikkinen 2009). It consists of a 20 cm-long and 1 cm-wide hand-held brass tube, which is mounted on a universal joint to make it hang vertically during operation. The vertical tube (or Cajanus tube) has not been widely used, but Johansson (1985) reported that the instrument is a reliable method for the measurement of crown projection, provided that the number of points sampled is adequate.

A Moosehorn is an instrument for measuring crown cover that was developed by Robinson (1947), and which was later redesigned by Garrison (1949). The instrument uses 25 dots in a 5 × 5 square array painted on a transparent screen and was designed to be very simple to use. Correct use of the instrument requires several steps: (i) holding the instrument in a vertical position, (ii) looking through a side aperture, and (iii) counting the number of dots that are covered by tree crowns. According to Fiala et al. (2006), the instrument is particularly well suited for rapid, efficient estimates of vertically projected canopy cover. However, Brown et al. (2000a) found the moosehorn to be cumbersome and fragile, and ended up redesigning it with several improvements.

The canopy-scope is an instrument that is used for the rapid assessment of canopy openness and it is relatively inexpensive, robust, accurate, and repeatable (Brown et al. 2000a). It is made of a transparent Perspex screen with a 20 cm cord attached to one corner. The cord allows the user to maintain the instrument at the same distance from the eye. As with the moosehorn, the screen of the canopy-scope is engraved with 25 dots in a 5 × 5 square array. The instrument is portable, robust, inexpensive to construct, and small enough to fit in a jacket pocket. Eight to ten measurements in a 0.25 ha plot seem to be sufficient for representing average canopy openness (Hale and Brown 2005). In comparison to other instruments, the canopy-scope was found to have lower within—and between-observer error than several other methods; it also had the highest correlation with canopy openness that was measured using hemispherical photographs (Brown et al. 2000a). For rapid assessments of canopy openness, the canopy-scope is the best instrument available when considering its low price, robustness, speed, and accuracy (Brown et al. 2000a).

The crown illumination index is a method for visually assessing the crown illumination of trees and was devised by Dawkins in 1958. Clark and Clark (1992) later modified the crown illumination index by subdividing class 2 (lateral light) into high, medium, and low lateral light. This widely used index allows the user to score the source and relative amount of light reaching a tree crown. It covers trees that do not receive any direct light either vertically or laterally (class 1) to trees that have completely exposed crowns (class 5; Clark and Clark 1992, p. 318). To overcome some of the limitations that are associated with using subjective indices such as the crown illumination index, a method based on crown illumination ellipses was developed. In this approach, a hole in the canopy is assessed by comparing it with a series of ellipses printed on a transparent Perspex screen (Brown et al. 2000a). As is the case with the canopy-scope, the instrument is held away from the observer with a 20 cm cord that is attached to the screen; the canopy cover can then be determined, based on the size of the ellipse that fits into the largest canopy opening that can be found in the canopy.

Single-point and handheld light sensors (e.g., AccuPAR, Decagon Devices, Pullman, WA, or Line Quantum Sensor, LI-COR Biosciences, Lincoln, NE) have long been used to estimate light transmittance through forest canopies (Pearcy 1989; Comeau 2000). Relative light transmittance can be calculated using two sensors. One is installed on a tower or in a large opening to record open sky light levels using a datalogger, and a second is used for light measurements in the understory. These measurements require clear-sky or completely overcast conditions, and variable conditions should be avoided.

Although the instruments that are described above for measuring canopy cover or canopy openness are convenient in the sense that they are simpler to use and technically less sophisticated than the equipment used in HP, they also have some drawbacks. For example, according to Brown et al. (2000a), resolution decreases for the crown illumination index, crown illumination ellipses, moosehorn, and canopy-scope when canopy openness exceeds 30%. Also, many of the aforementioned instruments have been shown to have noticeable within—and between-observer errors, although the canopy-scope is less prone to error and correlates well with canopy openness that is measured with HP (Brown et al. 2000a). Conversely, light measuring instruments such as single-point and handheld light sensors, while performing similarly to HP in various light environments (Gendron et al. 1998), only measure light transmittance; as such, they are less versatile than HP for characterizing canopy structure in forested environments.