Build the IPCC 2019 wet/dry drivers of the nitrogen loss cascade.

Put the IPCC 2019 Tier 1 wet/dry class of build_cell_moisture_regime(), and the Atlantic/Mediterranean drivers mapped from it, on every row of a gridded nitrogen balance, as the two driver tables build_nitrogen_balance() joins: n_balance_drivers (per balance key and loss fert_type, read by calculate_nh3(), calculate_soil_n2o() and calculate_indirect_n2o_nh3()) and n_balance_leaching_drivers (per balance key, read by calculate_n_leaching()). build_nitrogen_balance() calls it itself at grid resolution when neither table is supplied.

A row whose cell, polity, commodity balance item and water regime has a class of its own takes it. A row without one takes its cell's class on precipitation alone (no crop, no irrigation): a row with no crop item, as the soil organic matter sequestration rows, and a row of an item WHEP holds no land of in that water regime, as grass. A cell lacking CRU climate variables takes them from its nearest CRU land cell (see build_cell_moisture_regime()). Every row ends with a class.

climate is "ATL" on wet rows and "MED" on dry rows. irrig_type, the n2o_efs_disaggregated key of the Atlantic/Mediterranean options, is "Flooded" on every rice row (the commodity balance item of the production items classed Crop_irrig == "Rice" in items_prod_full) in both climates and both water regimes, "Med_average" on other irrigated MED rows, "Rainfed" on other rainfed MED rows and "Tier_1" on every other ATL row (the ATL table defines only "Tier_1" and "Flooded").

Usage

build_n_loss_drivers(
  rows = NULL,
  window = c("trailing_30", "annual", "normal"),
  water = c("precip_irrig", "precip"),
  data = NULL,
  config = list(),
  example = FALSE
)

Source

IPCC (2019), 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 4, Chapter 11, Table 11.1, footnote 4 (p. 11.13).

Arguments

rows: The balance rows to give drivers: lon, lat, area_code, item_cbs_code (NA on rows without a crop), year and water_regime ("rainfed" or "irrigated"), plus optionally fert_type, the loss fert_type of each row. With fert_type, the loss drivers are built for its non-missing (key, fert_type) pairs; without it, for every key and the seven loss fert_types ("Excreta", "Liquid", "Solid", "SOM", "Synthetic", "Human", "Recycling"). The leaching drivers cover every key. n_input_t, the N of each row, is needed wherever an irrigated row has no class of its own: such a row must carry no N.
window, water, config: As in build_cell_moisture_regime(). The window also sets the precipitation-only class of rows without a crop.
data: Optional list of inputs, passed to build_cell_moisture_regime() (climate, crop_depth, lpjml_years, lpjml_grid, regime_areas, country_grid), plus moisture_regime, a build_cell_moisture_regime() result to use instead of building it. Climate is read once and shared.
example: If TRUE, return a small fixture instead of reading data. Defaults to FALSE.

Value

A named list of two tibbles:

n_balance_drivers: one row per balance key (lon, lat, area_code, item_cbs_code, year, water_regime) and fert_type, with moisture_regime ("wet" or "dry"), climate ("ATL" or "MED"), irrig_type, method_moisture_regime, method_climate ("from_moisture_regime") and climate_fill_km.
n_balance_leaching_drivers: one row per balance key, with moisture_regime, method_moisture_regime and climate_fill_km.

Both are join tables keyed on the balance key; the balance they are joined to carries the polity columns.

Details

Which driver columns each loss method reads, and which this function supplies:

Method	Driver columns	Supplied
`nh3`: `"ipcc"`	`fert_type`	yes
`nh3`: `"manner"`, `"manner_default"`	MANNER's, see below	no
`n2o`: `"ipcc2019"`	`moisture_regime`	yes
`n2o`: `"ipcc2019_atl_med"`	`climate`	yes
`n2o`: `"aguilera"`, `"ipcc2006"`	`climate`, `irrig_type`	yes
`n2o_indirect`: `"ipcc2019"`	`moisture_regime`	yes
`n2o_indirect`: `"ipcc2019_atl_med"`	`climate`, `irrig_type`	yes
`leaching`: `"ipcc_fracleach"`	`moisture_regime`, `water_regime`	yes
`leaching`: `"meisinger_drainage"`	Meisinger's, see below	no

The two methods it does not supply are later work. MANNER reads manner_fertiliser, soil_ph, rate_kg_ha, rainfall_mm, irrigated, temp_c, temp_c_annual_mean, windspeed_ms, technique, system, incorporation_delay_h and species (see calculate_nh3()); Meisinger reads fert_type, climate, irrig_cat, land_use, cn_input, tillage, som_share and drainage_mm (see calculate_n_leaching()). build_nitrogen_balance() refuses either before assembling any input unless its driver table is supplied.

method_moisture_regime is build_cell_moisture_regime()'s stamp. A row that took its cell's precipitation-only class carries the extra field cell_rainfed_class=no_crop_item (no crop item) or cell_rainfed_class=no_regime_area (an item without regime land of the row's water regime), and a row in a cell that took climate variables from its nearest CRU cell the field nearest_cru_cell=<col>[+<col>...], naming them.

Errors

whep_n_loss_drivers_rows: rows is missing, empty, lacks a key column, or holds a missing coordinate or year, or a water_regime other than "rainfed" or "irrigated".
whep_n_loss_drivers_unclassified: a row is still without a class (for example a window-mean PET of zero), named by its stamp.
whep_n_loss_drivers_irrigated_cell_class: an irrigated row carrying N, or one whose N is not supplied, has no class of its own and would take its cell's precipitation-only class. The rainfed/irrigated split gives an irrigated row N only where the item holds irrigated land, so such a row means the balance and the moisture regime disagree on the item's regime land; the error names the count and the items carrying most N.
Every error of build_cell_moisture_regime(), including whep_absent_input for an absent climate or irrigation layer.

Examples

build_n_loss_drivers(example = TRUE)
#> $n_balance_drivers
#> # A tibble: 5 × 13
#>     lon   lat area_code item_cbs_code  year water_regime fert_type
#>   <dbl> <dbl>     <int>         <int> <int> <chr>        <chr>    
#> 1  9.25  47.8        79          2514  2010 irrigated    Synthetic
#> 2  9.25  47.8        79            NA  2010 rainfed      Synthetic
#> 3  9.75  47.8        79          3000  2010 rainfed      Synthetic
#> 4  9.75  47.8        79          2807  2010 irrigated    Synthetic
#> 5 10.2   47.8        79          2514  2010 rainfed      Synthetic
#> # ℹ 6 more variables: moisture_regime <chr>, climate <chr>, irrig_type <chr>,
#> #   method_moisture_regime <chr>, method_climate <chr>, climate_fill_km <dbl>
#> 
#> $n_balance_leaching_drivers
#> # A tibble: 5 × 9
#>     lon   lat area_code item_cbs_code  year water_regime moisture_regime
#>   <dbl> <dbl>     <int>         <int> <int> <chr>        <chr>          
#> 1  9.25  47.8        79          2514  2010 irrigated    dry            
#> 2  9.25  47.8        79            NA  2010 rainfed      dry            
#> 3  9.75  47.8        79          3000  2010 rainfed      wet            
#> 4  9.75  47.8        79          2807  2010 irrigated    wet            
#> 5 10.2   47.8        79          2514  2010 rainfed      wet            
#> # ℹ 2 more variables: method_moisture_regime <chr>, climate_fill_km <dbl>
#>