dwm: Estimate volume, biomass, and carbon stocks of down woody...
In rFIA: Estimation of Forest Variables using the FIA Database

Description Usage Arguments Details Value Note Author(s) References See Also Examples

Produces estimates of down woody material stocks on a per acre basis from the Forest Inventory and Analysis Database (FIADB), along with population totals for each variable. Estimates are returned by fuel class (duff, litter, 1HR, 10HR, 100HR, 1000HR, piles) for application in fuels management. Estimates can be produced for regions defined within the FIA Database (e.g. counties), at the plot level, or within user-defined areal units. If multiple reporting years (EVALIDs) are included in the data, estimates will be output as a time series. If multiple states are represented by the data, estimates will be output for the full region (all area combined), unless specified otherwise (e.g. grpBy = STATECD).

dwm(db, grpBy = NULL, polys = NULL, returnSpatial = FALSE, landType = 'forest',
    method = 'TI', lambda = .5, areaDomain = NULL, totals = FALSE,
    variance = FALSE, byPlot = FALSE, condList = FALSE,
    byFuelType = TRUE, nCores = 1)

`db`	`FIA.Database` or `Remote.FIA.Database` object produced from `readFIA` or `getFIA`. If a `Remote.FIA.Database`, data will be read in and processed state-by-state to conserve RAM (see details for an example).
`grpBy`	variables from PLOT, COND, or TREE tables to group estimates by (NOT quoted). Multiple grouping variables should be combined with `c()`, and grouping will occur heirarchically. For example, to produce seperate estimates for each ownership group within ecoregion subsections, specify `c(ECOSUBCD, OWNGRPCD)`.
`polys`	`sp` or `sf` Polygon/MultiPolgyon object; Areal units to bin data for estimation. Seperate estimates will be produces for region encompassed by each areal unit. FIA plot locations will be reprojected to match projection of `polys` object.
`returnSpatial`	logical; if TRUE, merge population estimates with `polys` and return as `sf` multipolygon object. When `byPlot = TRUE`, return plot-level estimates as `sf` spatial points.
`landType`	character ("forest" or "timber""); Type of land that estimates will be produced for. Timberland is a subset of forestland (default) which has high site potential and non-reserve status (see details).
`method`	character; design-based estimator to use. One of: "TI" (temporally indifferent, default), "annual" (annual), "SMA"" (simple moving average), "LMA" (linear moving average), or "EMA" (exponential moving average). See Stanke et al 2020 for a complete description of these estimators.
`lambda`	numeric (0,1); if `method = 'EMA'`, the decay parameter used to define weighting scheme for annual panels. Low values place higher weight on more recent panels, and vice versa. Specify a vector of values to compute estimates using mulitple wieghting schemes, and use `plotFIA` with `grp` set to `lambda` to produce moving average ribbon plots. See Stanke et al 2020 for examples.
`areaDomain`	Logical predicates defined in terms of the variables in PLOT and/or COND tables. Used to define the area for which estimates will be produced (e.g. within 1 mile of improved road: `RDDISTCD %in% c(1:6)`, Hard maple/basswood forest type: `FORTYPCD == 805)`. Multiple conditions are combined with `&` (and) or `\|` (or). Only plots within areas where the condition evaluates to TRUE are used in producing estimates. Should NOT be quoted.
`totals`	logical; if TRUE, return total population estimates (e.g. total area) along with ratio estimates (e.g. mean trees per acre).
`variance`	logical; if TRUE, return estimated variance (`VAR`) and sample size (`N`). If FALSE, return 'sampling error' (`SE`) as returned by EVALIDator. Note: sampling error cannot be used to construct confidence intervals.
`byPlot`	logical; if TRUE, returns estimates for individual plot locations instead of population estimates.
`condList`	logical; if TRUE, returns condition-level summaries intended for subsequent use with `customPSE`.
`byFuelType`	logical; if TRUE, returns estimates grouped by fuel type (e.g., 1HR, 10HR, 100HR, 1000HR fuels).
`nCores`	numeric; number of cores to use for parallel implementation. Check available cores using `detectCores`. Default = 1, serial processing.

Estimation Details

Estimation of forest variables follows the procedures documented in Bechtold and Patterson (2005) and Stanke et al 2020. Specifically, per acre estimates are computed using a sample-based ratio-of-means estimator of total volume (biomass or carbon) / total land area within the domain of interest.

As defined by FIA, down woody material includes dead organic materials (resulting from plant mortality and leaf turnover) and fuel complexes of live shrubs and herbs. To maintain relevance for forest fuels management, we report estimates grouped by fuel lag-time classes. Specifically, we report estimates for 1HR fuels (small, fine woody debris), 10HR fuels (medium, fine woody debris), 100HR fuels (large, fine woody debris), 1000HR fuels (coarse woody debris), and slash piles, in addition to duff (O horizon; all unidentifiable organic material above mineral soil, beneath litter) and litter (identifiable plant material which is downed and smaller than 10HR fuel class (1HR class includes standing herbaceous material). See Woodall and Monleon (2007) for definitions of fuel lag-time classes and for details on sampling and estimation procedures.

Users may specify alternatives to the 'Temporally Indifferent' estimator using the method argument. Alternative design-based estimators include the annual estimator ("ANNUAL"; annual panels, or estimates from plots measured in the same year), simple moving average ("SMA"; combines annual panels with equal weight), linear moving average ("LMA"; combine annual panels with weights that decay linearly with time since measurement), and exponential moving average ("EMA"; combine annual panels with weights that decay exponentially with time since measurement). The "best" estimator depends entirely on user-objectives, see Stanke et al 2020 for a complete description of these estimators and tradeoffs between precision and temporal specificity.

When byPlot = FALSE (i.e., population estimates are returned), the "YEAR" column in the resulting dataframe indicates the final year of the inventory cycle that estimates are produced for. For example, an estimate of current forest area (e.g., 2018) may draw on data collected from 2008-2018, and "YEAR" will be listed as 2018 (consistent with EVALIDator). However, when byPlot = TRUE (i.e., plot-level estimates returned), the "YEAR" column denotes the year that each plot was measured (MEASYEAR), which may differ slightly from its associated inventory year (INVYR).

Stratified random sampling techniques are most often employed to compute estimates in recent inventories, although double sampling and simple random sampling may be employed for early inventories. Estimates are adjusted for non-response bias by assuming attributes of non-response plot locations to be equal to the mean of other plots included within thier respective stratum or population.

Working with "Big Data"

If FIA data are too large to hold in memory (e.g., R throws the "cannot allocate vector of size ..." errors), use larger-than-RAM options. See documentation of link{readFIA} for examples of how to set up a Remote.FIA.Database. As a reference, we have used rFIA's larger-than-RAM methods to estimate forest variables using the entire FIA Database (~50GB) on a standard desktop computer with 16GB of RAM. Check out our website for more details and examples.

Easy, efficient parallelization is implemented with the parallel package. Users must only specify the nCores argument with a value greater than 1 in order to implement parallel processing on their machines. Parallel implementation is achieved using a snow type cluster on any Windows OS, and with multicore forking on any Unix OS (Linux, Mac). Implementing parallel processing may substantially decrease free memory during processing, particularly on Windows OS. Thus, users should be cautious when running in parallel, and consider implementing serial processing for this task if computational resources are limited (nCores = 1).

Definition of forestland

Forest land must be at least 10-percent stocked by trees of any size, including land that formerly had such tree cover and that will be naturally or artificially regenerated. Forest land includes transition zones, such as areas between heavily forested and nonforested lands that are at least 10-percent stocked with trees and forest areas adjacent to urban and builtup lands. The minimum area for classification of forest land is 1 acre and 120 feet wide measured stem-to-stem from the outer-most edge. Unimproved roads and trails, streams, and clearings in forest areas are classified as forest if less than 120 feet wide. Timber land is a subset of forest land that is producing or is capable of producing crops of industrial wood and not withdrawn from timber utilization by statute or administrative regulation. (Note: Areas qualifying as timberland are capable of producing at least 20 cubic feet per acre per year of industrial wood in natural stands. Currently inaccessible and inoperable areas are NOT included).

Dataframe or SF object (if returnSpatial = TRUE). If byPlot = TRUE, values are returned for each plot (PLOT_STATUS_CD = 1 when forest exists at the plot location). All variables with names ending in SE, represent the estimate of sampling error (%) of the variable. When variance = TRUE, variables ending in VAR denote the variance of the variable and N is the total sample size (i.e., including non-zero plots).

YEAR: reporting year associated with estimates
VOL_ACRE: estimate of mean volume per acre of dwm (cu.ft/acre)
BIO_ACRE: estimate of mean biomass per acre of dwm (short tons/acre)
CARB_ACRE: estimate of mean carbon mass per acre of dwm (short tons/acre)
nPlots: number of non-zero plots used to compute estimates

All sampling error estimates (SE) are returned as the "percent coefficient of variation" (standard deviation / mean * 100) for consistency with EVALIDator. IMPORTANT: sampling error cannot be used to construct confidence intervals. Please use variance = TRUE for that (i.e., return variance and sample size instead of sampling error).

Hunter Stanke and Andrew Finley

rFIA website: https://rfia.netlify.app/

FIA Database User Guide: https://www.fia.fs.fed.us/library/database-documentation/

Bechtold, W.A.; Patterson, P.L., eds. 2005. The Enhanced Forest Inventory and Analysis Program - National Sampling Design and Estimation Procedures. Gen. Tech. Rep. SRS - 80. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southern Research Station. 85 p. https://www.srs.fs.usda.gov/pubs/gtr/gtr_srs080/gtr_srs080.pdf

Stanke, H., Finley, A. O., Weed, A. S., Walters, B. F., & Domke, G. M. (2020). rFIA: An R package for estimation of forest attributes with the US Forest Inventory and Analysis database. Environmental Modelling & Software, 127, 104664.

Woodall, C.; Monleon, V.J., eds. 2007. Sampling Protocol, Estimation, and Analysis Procedures for the Down Woody Materials Indicator of the FIA Program. Gen. Tech. Rep. NRS - 22. ewtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. https://www.nrs.fs.fed.us/pubs/gtr/gtr_nrs22.pdf

tpa, biomass

## Load data from rFIA package
data(fiaRI)
data(countiesRI)

## Most recents subset
fiaRI_mr <- clipFIA(fiaRI)

## Most recent estimates
dwm(fiaRI_mr)


## Same as above at the plot-level
## Most recent estimates
dwm(fiaRI_mr, byPlot = TRUE)

## Estimates of all forestland, over time
dwm(fiaRI)

## Estimates of all forestland on mesic sites (most recent)
dwm(fiaRI_mr,
    areaDomain = PHYSCLCD %in% 21:29)

## Estimates of all forestland by owner group (most recent subset)
dwm(fiaRI_mr,
    grpBy = OWNGRPCD)

## Estimates of all forestland by county and return
## return spatial object
dwmSF <- dwm(fiaRI_mr,
             polys = countiesRI,
             returnSpatial = TRUE)
plot(dwmSF)
plotFIA(dwmSF, BIO_ACRE) # TOTAL BIOMASS / ACRE (tons)