nstdmetrics: Predefined non standard metrics

In lidR: Airborne LiDAR Data Manipulation and Visualization for Forestry Applications

Predefined non standard metrics

Description

Functions and metrics from the literature. See details and references

Usage

rumple_index(x, y = NULL, z = NULL, ...)

gap_fraction_profile(z, dz = 1, z0 = 2)

LAD(z, dz = 1, k = 0.5, z0 = 2)

entropy(z, by = 1, zmax = NULL)

VCI(z, zmax, by = 1)

Arguments

x	A RasterLayer, a stars a SpatRaster or a vector of x coordinates.
y	numeric. If x is a vector of coordinates: the associated y coordinates.
z	vector of positive z coordinates
...	unused
z0	numeric. The bottom limit of the profile
k	numeric. is the extinction coefficient
by, dz	numeric. The thickness of the layers used (height bin)
zmax	numeric. Maximum elevation for an entropy normalized to zmax.

Details

rumple_index

Computes the roughness of a surface as the ratio between its area and its projected area on the ground. If the input is a gridded object (raster) the function computes the surfaces using Jenness's algorithm (see references). If the input is a point cloud the function uses a Delaunay triangulation of the points and computes the area of each triangle.

gap_fraction_profile

Computes the gap fraction profile using the method of Bouvier et al. (see reference). The function assesses the number of laser points that actually reached the layer z+dz and those that passed through the layer [z, z+dz]. By definition the layer 0 will always return 0 because no returns pass through the ground. Therefore, the layer 0 is removed from the returned results.

LAD

Computes a leaf area density profile based on the method of Bouvier et al. (see reference) The function assesses the number of laser points that actually reached the layer z+dz and those that passed through the layer [z, z+dz] (see gap_fraction_profile). Then it computes the log of this quantity and divides it by the extinction coefficient k as described in Bouvier et al. By definition the layer 0 will always return infinity because no returns pass through the ground. Therefore, the layer 0 is removed from the returned results.

entropy

A normalized Shannon vertical complexity index. The Shannon diversity index is a measure for quantifying diversity and is based on the number and frequency of species present. This index, developed by Shannon and Weaver for use in information theory, was successfully transferred to the description of species diversity in biological systems (Shannon 1948). Here it is applied to quantify the diversity and the evenness of an elevational distribution of las points. It makes bins between 0 and the maximum elevation. If there are negative values the function returns NA.

VCI

Vertical Complexity Index. A fixed normalization of the entropy function from van Ewijk et al. (2011) (see references)

Value

numeric. The computed Rumple index.

A data.frame containing the bin elevations (z) and the gap fraction for each bin (gf)

A number between 0 and 1

A number between 0 and 1

References

Jenness, J. S. (2004). Calculating landscape surface area from digital elevation models. Wildlife Society Bulletin, 32(3), 829–839.

Bouvier, M., Durrieu, S., Fournier, R. a, & Renaud, J. (2015). Generalizing predictive models of forest inventory attributes using an area-based approach with airborne las data. Remote Sensing of Environment, 156, 322-334. http://doi.org/10.1016/j.rse.2014.10.004

Pretzsch, H. (2008). Description and Analysis of Stand Structures. Springer Berlin Heidelberg. http://doi.org/10.1007/978-3-540-88307-4 (pages 279-280) Shannon, Claude E. (1948), "A mathematical theory of communication," Bell System Tech. Journal 27, 379-423, 623-656.

van Ewijk, K. Y., Treitz, P. M., & Scott, N. A. (2011). Characterizing Forest Succession in Central Ontario using LAS-derived Indices. Photogrammetric Engineering and Remote Sensing, 77(3), 261-269. Retrieved from <Go to ISI>://WOS:000288052100009

Examples

x <- runif(20, 0, 100)
y <- runif(20, 0, 100)

if (require(geometry, quietly = TRUE))
{
# Perfectly flat surface, rumple_index = 1
z <- rep(10, 20)
rumple_index(x, y, z)

# Rough surface, rumple_index > 1
z <- runif(20, 0, 10)
rumple_index(x, y, z)

# Rougher surface, rumple_index increases
z <- runif(20, 0, 50)
rumple_index(x, y, z)

# Measure of roughness is scale-dependent
rumple_index(x, y, z)
rumple_index(x/10, y/10, z)
}
z <- c(rnorm(1e4, 25, 6), rgamma(1e3, 1, 8)*6, rgamma(5e2, 5,5)*10)
z <- z[z<45 & z>0]

hist(z, n=50)

gapFraction = gap_fraction_profile(z)

plot(gapFraction, type="l", xlab="Elevation", ylab="Gap fraction")
z <- c(rnorm(1e4, 25, 6), rgamma(1e3, 1, 8)*6, rgamma(5e2, 5,5)*10)
z <- z[z<45 & z>0]

lad <- LAD(z)

plot(lad, type="l", xlab="Elevation", ylab="Leaf area density")
z <- runif(10000, 0, 10)

# expected to be close to 1. The highest diversity is given for a uniform distribution
entropy(z, by = 1)

 z <- runif(10000, 9, 10)

# Must be 0. The lowest diversity is given for a unique possibility
entropy(z, by = 1)

z <- abs(rnorm(10000, 10, 1))

# expected to be between 0 and 1.
entropy(z, by = 1)
z <- runif(10000, 0, 10)

VCI(z, by = 1, zmax = 20)

z <- abs(rnorm(10000, 10, 1))

# expected to be closer to 0.
VCI(z, by = 1, zmax = 20)

lidR documentation built on Sept. 11, 2024, 5:21 p.m.