This function attributes a weight to each equation based on its sampling size, and taxonomic and climatic similarity with the species/site combination considered.
Arguments
- genus
a character value, containing the genus (e.g. "Quercus") of the tree.
- coords
a numeric vector of length 2 with longitude and latitude.
- species
a character vector (same length as genus), containing the species (e.g. "rubra") of the tree. Default is
NULL
, when no species identification is available.- new_eqtable
Optional. An equation table created with the
new_equations()
function.- wna
a numeric vector, this parameter is used in the
weight_allom()
function to determine the sample-size related weights attributed to equations without a specified sample size. Default is 0.1.- w95
a numeric vector, this parameter is used to determine the value at which the sample-size-related weight reaches 95% of its maximum value (max=1). Default is 500.
Details
Each equation is given a weight by the function weight_allom()
, calculated
as the product of the following components:
(1) sample-size weight, calculated as:
$$1-exp(-n*(log(20)/w95))$$
where n is the sample size of the equation; the weight given to equations
with no sample size information is determined by argument wna
(0.1 by
default).
(2) climate weight, based on the similarity between the climatic conditions
of the equation site and the target location, using the three-letter system
of Koppen's climate scheme. Climate weights associated with each combination
of two Koppen climates are provided in data("koppenMatrix")
. The resulting
weight has a value between 1e-6 (different climate groups) and 1 (exactly the
same climate classification). When an equation was calibrated with trees from
several locations with different Koppen climates, the maximum value out of
all pairwise equation-site climate weights is used.
(3) taxonomic weight: equal to 1 for same species equations, 0.8 for same genus equations, 0.5 for same family equations and for equations calibrated for the same broad functional or taxonomic group (e.g. shrubs, conifers, angiosperms). All other equations are given a low taxonomic weight of 1e-6: these equations will have a significant relative weight in the final prediction only when no other more specific equation is available.
Examples
x <- weight_allom(
genus = "Acer",
species = "negundo",
coords = c(-78.2, 38.9)
)
str(x)
#> Named num [1:477] 6.13e-08 5.25e-08 1.00e-07 1.00e-07 1.00e-07 ...
#> - attr(*, "names")= chr [1:477] "4b4063" "e2c7c7" "e42e41" "9f138f" ...
head(x)
#> 4b4063 e2c7c7 e42e41 9f138f 9e3b6a e57b77
#> 6.134069e-08 5.245305e-08 1.000000e-07 1.000000e-07 1.000000e-07 1.000000e-07