Skip to content

Down Woody Material

Down woody material (DWM) refers to a set of attributes involving the volume, biomass, and other characteristics of detritus on the forest floor. DWM analysis involves particular field methods including line transect sampling, among others, that require special consideration when creating plot-level values. Currently, fiaplyr supports DWM analysis up to, but no further than, a pre-compiled DWM table referred to as COND_DWM_CALC. This table is prepared by FIA, and standardizes the DWM attributes for each condition in the FIA database. However, use of this table is fairly restrictive, preventing e.g., piece-level domains for coarse woody debris and other creative ideas. Still, basic analysis of DWM is possible, and we demonstrate this with an example from Vermont.

Much like the GRM case, users produce estimates using a series of facilitating functions called targets, which are called within aggregate and estimate contexts. The target names should be familiar, and are all prefixed with dwm_. Common examples include:

  • dwm_cwd
  • dwm_fwd
  • dwm_litter
  • dwm_pile

These targets correspond to different components of DWM analysis, and the supported variables differ by component, as shown in the following table:

ComponentTargetSupported variables
Coarse woody debrisdwm_cwd()VOLCF, DRYBIO, CARBON, LPA
Fine woody debrisdwm_fwd()VOLCF, DRYBIO, CARBON
Residual pilesdwm_pile()VOLCF, DRYBIO, CARBON
Fuelsdwm_fuel()DRYBIO, CARBON
Duffdwm_duff()DRYBIO, CARBON
Litterdwm_litter()DRYBIO, CARBON

Units are cubic feet per acre for VOLCF, dry short tons per acre for DRYBIO, short tons per acre for CARBON, and pieces per acre for LPA. dwm_fwd() additionally requires a size argument of "SM", "MD", "LG", or "ALL" (the last summing all three classes).

To begin with DWM analyses, specify a handler as always. However, this handler must use an evaluation ending in 7, which indicates a DWM-specific evaluation. We also pass the dwm_analysis() specification.

library(fiaplyr)
library(DBI)
library(duckdb)
library(dplyr)
con <- dbConnect(duckdb(), fiadb_vt_mini_path())
handler <- eval_handler(con, 501007, spec = dwm_analysis())
handler
EvalHandler
----------
EVALID: 501007
Description: VERMONT 2010: 2006-2010: DWM
Plots: 58
Inventory Years: 2006 - 2010
Measure Years: 2006 - 2010

Here, we see the handler is initialized with a DWM evaluation, and the summary statement gives particulars.

Estimation of Down Woody Material Components

Section titled “Estimation of Down Woody Material Components”

Beginning with a simple example, let’s assume we want to estimate the total coarse woody debris biomass. The target dwm_cwd is used, which can accept the DRYBIO variable, yielding

handler |>
estimate(
dwm(
cwd_bio = dwm_cwd(DRYBIO)
),
output = "total"
) |>
collect()
# A tibble: 1 × 3
var estimate se
<chr> <dbl> <dbl>
1 cwd_bio 24318923. 4263451.

Likewise, the biomass of other components can be estimated with the appropriate targets, such as dwm_fwd, which accepts a size class argument

handler |>
estimate(
dwm(
fwd_sm_bio = dwm_fwd(DRYBIO, size = "SM")
),
output = "total"
) |>
collect()
# A tibble: 1 × 3
var estimate se
<chr> <dbl> <dbl>
1 fwd_sm_bio 342676. 36070.

Before estimation, plot-level aggregates can be obtained with aggregate(). Unlike estimate(), aggregation operates on the _UNADJ DWM fields and returns one row per plot:

handler |>
aggregate(dwm(dwm_cwd(VOLCF))) |>
collect() |>
head()
# A tibble: 6 × 6
PLT_CN STATECD COUNTYCD INVYR PLOT dwm_cwd_VOLCF
<chr> <int> <int> <int> <int> <dbl>
1 118137344010661 50 25 2008 1130 141.
2 118137372010661 50 27 2008 635 231.
3 118137390010661 50 3 2008 405 737.
4 118137426010661 50 5 2008 1129 665.
5 118137460010661 50 17 2008 33 341.
6 118137512010661 50 19 2008 1484 0

Like estimate(), aggregate() returns a lazy query; collect() is used above to bring the result into memory.

Estimation of Ratios Involving Down Woody Material

Section titled “Estimation of Ratios Involving Down Woody Material”

DWM estimates often need to be expressed as ratios, especially when normalizing to forested area. Estimates of this kind proceed along the same lines as ratio estimates. The only modification, of course, is the specification of the DWM estimates themselves.

handler |>
subset(cond(COND_STATUS_CD == 1)) |>
estimate(
ratio(
dwm(dwm_cwd(CARBON)),
cond()
)
) |>
collect()
# A tibble: 1 × 4
var_n var_d estimate se
<chr> <chr> <dbl> <dbl>
1 dwm_cwd_CARBON prop 2.72 0.461