Skip to contents

Decompose semi-natural agroecosystem N surplus into size, intensity, and inefficiency drivers

Source: R/decomposition_analysis.R
decompose_semi_natural_surplus.Rd

Decomposes the year-on-year change in Spain's semi-natural agroecosystem (grazing land, dehesa, and non-cropland vegetation) nitrogen (N) surplus into three multiplicative drivers, following an additive LMDI shift-share decomposition computed at the national level:

  • Size: national semi-natural area.

  • Intensity: N input per hectare of semi-natural land.

  • Inefficiency: surplus fraction of inputs (1 - nitrogen use efficiency).

No destiny factor is used because grazed and cut vegetation is assumed to be overwhelmingly a single destiny (livestock feed).

The land-use categories included (Dehesa, Forest_high, Forest_low, Other, Pasture_Shrubland) are all of npp_ygpit's non-cropland categories, matching the existing semi_natural_agroecosystems box used elsewhere in the package. Some of that land (e.g. Forest_high/ Forest_low) may not actually be grazed and can produce non-feed output (firewood), which would call for its own destiny factor (as in decompose_cropland_surplus()) rather than the single-destiny assumption above; that refinement is not implemented here.

This is a simplified, national-only view (no provincial breakdown). Semi-natural surplus can turn negative (soil N mining) in some years. LMDI relies on logarithms and cannot handle a series that changes sign between two compared years; this function warns when that occurs instead of silently returning NA, but does not implement the Shapley/Sun alternative required for those cases.

Usage

decompose_semi_natural_surplus(
  n_prov_destiny = NULL,
  npp_ygpit = NULL,
  by_period = FALSE,
  example = FALSE
)

Arguments

n_prov_destiny

Nitrogen flows tibble from create_n_prov_destiny(). If NULL, loaded automatically.

npp_ygpit

Land use and area tibble from whep_read_file("npp_ygpit"). If NULL, loaded automatically.

by_period

If TRUE, compares each reference period (each averaged across its ten years) against the immediately preceding one — 1860-1870 -> 1920-1930 -> 1960-1970 -> 2010-2020 — plus one extra transition spanning the full analysis window, 1860-1870 straight to 2010-2020 (the total change) — instead of chaining year on year.

example

If TRUE, return a small hardcoded output without downloading remote data. Default is FALSE.

Value

A tibble from calculate_lmdi() with columns period, period_years, factor_label, component_type, additive, multiplicative, and multiplicative_log.

Examples

decompose_semi_natural_surplus(example = TRUE)
#> # A tibble: 8 × 7
#>   period    period_years factor_label     component_type additive multiplicative
#>   <chr>            <dbl> <chr>            <chr>             <dbl>          <dbl>
#> 1 1860-1861            1 Size             factor            -287.          0.997
#> 2 1860-1861            1 Intensity        factor           -3265.          0.965
#> 3 1860-1861            1 Inefficiency     factor          -10384.          0.892
#> 4 1860-1861            1 Semi-natural N … target          -13936.          0.858
#> 5 1861-1862            1 Size             factor            -401.          0.997
#> 6 1861-1862            1 Intensity        factor           30174.          1.27 
#> 7 1861-1862            1 Inefficiency     factor           69339.          1.72 
#> 8 1861-1862            1 Semi-natural N … target           99112.          2.17 
#> # ℹ 1 more variable: multiplicative_log <dbl>