R/zchunk_LB122.LC_R_Cropland_Yh_GLU.R
In JGCRI/gcamdata: Build the GCAM Input Datasets
Defines functions
module_aglu_LB122.LC_R_Cropland_Yh_GLU
Documented in
module_aglu_LB122.LC_R_Cropland_Yh_GLU
# Copyright 2019 Battelle Memorial Institute; see the LICENSE file.
#' module_aglu_LB122.LC_R_Cropland_Yh_GLU
#'
#' Integrate disparate data sources for land cover and harvested area from FAO, Monfreda, and Hyde in
#' order to calculate cropland cover by specific crop type, other arable land, harvested:cropped ratio, economic
#' yield (i.e., production per unit cropland area per year, rather than production per unit area per harvest), and
#' "extra" cropland that needs to be taken from other land use types. All data tables are written out
#' at the GCAM region-GLU level for each historical year, including carbon cycle model spinup years.
#'
#' @param command API command to execute
#' @param ... other optional parameters, depending on command
#' @return Depends on \code{command}: either a vector of required inputs,
#' a vector of output names, or (if \code{command} is "MAKE") all
#' the generated outputs: \code{L122.ag_HA_to_CropLand_R_Y_GLU}, \code{L122.ag_EcYield_kgm2_R_C_Y_GLU},
#' \code{L122.LC_bm2_R_OtherArableLand_Yh_GLU}, \code{L122.LC_bm2_R_ExtraCropLand_Yh_GLU},
#' \code{L122.LC_bm2_R_HarvCropLand_C_Yh_GLU}, \code{L122.LC_bm2_R_HarvCropLand_Yh_GLU}.
#' The corresponding file in the original data system was \code{LB122.LC_R_Cropland_Yh_GLU.R} (aglu level1).
#' @details First, Hyde cropland cover is aggregated to the region-GLU level in each year, and supplemented with Monfreda
#' harvested area information when necessary. Second, FAO fallowland, cropland, and harvested area are aggregated from
#' the country level to the region level, and used to calculate the "residual" cropland in each land use region:
#' total cropland (Hyde) - known fallow (FAO) - harvested crops (FAO/Monfreda).
#' The harvested area : cropland ratio (HA:CL) is computed as (sum of harvested area) / (cropland - fallow land).
#' Where HA:CL < 1, HA:CL is set to 1, and the balance (the "residual" above) is added to OtherArableLand.
#' Where HA:CL > 3, HA:CL is set to 3, and additional cropland (ExtraCropLand) is written out, to be taken from
#' other land use types (determined in different code chunks).
#' @importFrom assertthat assert_that
#' @importFrom dplyr anti_join arrange bind_rows distinct filter if_else group_by left_join mutate select summarise summarise_all
#' @importFrom tidyr replace_na
#' @author ACS April 2017
module_aglu_LB122.LC_R_Cropland_Yh_GLU <- function(command, ...) {
if(command == driver.DECLARE_INPUTS) {
return(c(FILE = "common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L101.ag_Prod_Mt_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU"))
} else if(command == driver.DECLARE_OUTPUTS) {
return(c("L122.ag_HA_to_CropLand_R_Y_GLU",
"L122.ag_EcYield_kgm2_R_C_Y_GLU",
"L122.LC_bm2_R_OtherArableLand_Yh_GLU",
"L122.LC_bm2_R_ExtraCropLand_Yh_GLU",
"L122.LC_bm2_R_HarvCropLand_C_Yh_GLU",
"L122.LC_bm2_R_HarvCropLand_Yh_GLU"))
} else if(command == driver.MAKE) {
Land_Type <- year <- . <- GCAM_region_ID <- GLU <- GCAM_commodity <-
value <- iso <- countries <- cropland <- fallow <- fallow_frac <- cropped <-
cropped_frac <- uncropped_frac <- nonharvested_frac <- value.x <-
value.y <- Land_Type.y <- Land_Type.x <- GCAM_subsector <- NULL # silence package check.
all_data <- list(...)[[1]]
# Load required inputs
iso_GCAM_regID <- get_data(all_data, "common/iso_GCAM_regID")
L100.FAO_fallowland_kha <- get_data(all_data, "L100.FAO_fallowland_kha")
L100.FAO_CL_kha <- get_data(all_data, "L100.FAO_CL_kha")
L100.FAO_harv_CL_kha <- get_data(all_data, "L100.FAO_harv_CL_kha")
L101.ag_HA_bm2_R_C_Y_GLU <- get_data(all_data, "L101.ag_HA_bm2_R_C_Y_GLU", strip_attributes = TRUE)
L101.ag_Prod_Mt_R_C_Y_GLU <- get_data(all_data, "L101.ag_Prod_Mt_R_C_Y_GLU")
L120.LC_bm2_R_LT_Yh_GLU <- get_data(all_data, "L120.LC_bm2_R_LT_Yh_GLU")
# Line 36 in original file
# take a subset of the land cover table L120.LC_bm2_R_LT_YH_GLU: cropland, and only in aglu historical years
L120.LC_bm2_R_LT_Yh_GLU %>%
filter(Land_Type == "Cropland", year %in% aglu.AGLU_HISTORICAL_YEARS) ->
# store in table Land Cover in bm2 by Region, Year and GLU for Cropland.
L122.LC_bm2_R_CropLand_Y_GLU
# Lines 40-45 in original file
# The harvested area/production tables L103.ag.X (from Monfreda) may have R_GLUs not in the cropland table
# L122.LC_bm2_R_CropLand_Y_GLU (from Hyde), and vice versa.
# Fill out the cropland table to include all R_GLUs in Monfreda:
L122.LC_bm2_R_CropLand_Y_GLU %>%
# find the region-GLU combos in the Monfreda harvested area table L101.ag_HA_bm2_R_C_Y_GLU NOT contained
# in the L122.LC_bm2_R_CropLand_Y_GLU land cover table:
anti_join(L101.ag_HA_bm2_R_C_Y_GLU[,c("GCAM_region_ID", "GLU", "year")], .,
by = c("GCAM_region_ID", "GLU", "year")) %>%
# save only the unique combinations:
unique() %>%
# arrange so rows occur in more sensible order:
arrange(GCAM_region_ID, GLU, year) %>%
# add values of 0 everywhere for now:
mutate(value = 0,
# add a Land_Type identifier:
Land_Type = "Cropland") %>%
# add these rows to the original Land Cover table:
bind_rows(L122.LC_bm2_R_CropLand_Y_GLU) ->
# save as the Land C over table:
L122.LC_bm2_R_CropLand_Y_GLU
# Line 48 in original file
# old comment: compile harvested area across all crops
# Take the input historical harvested area table, L101.ag_HA_bm2_R_C_Y_GLU, and sum over GCAM_commodity, so that each
# region-GLU-year combo has a single value.
L101.ag_HA_bm2_R_C_Y_GLU %>%
group_by(GCAM_region_ID, GLU, year) %>%
# and sum by region-GLU-year:
summarise(value = sum(value)) %>%
ungroup() ->
L122.ag_HA_bm2_R_Y_GLU
# Calculating the average percent of cropland that is fallow in each region
# using the average cropland, fallow land, and land in temporary crops from FAO RESOURCESTAT
# The time series is unreliable, so using the average of all available years between 2008 and 2012
# (and applying to all historical years)
# And continue with calculating average crop land, fallow land, land in temporary crops:
# compile the ratio of "temporary crops" to total arable land in each region
# make table with fallow land compared to total arable land
# FAO fallow land data, L100.FAO_fallowland_kha, is appended to the FAO cropland data table. Quantities are aggregated
# from iso to GCAM region and the fraction of fallowland is calculated.
#
# The old data sytem uses match to append FAO fallow land data to FAO cropland data by iso. However, Ethiopia,
# Sudan, and Belgium-Luxembourg have all split into two distinct countries and FAO produces data for each.
# Specifically, there are two rows with iso = "eth" and two rows with iso = "sdn" and two rows with iso = "bel" in some FAO data.
# Because match only pulls the first value when there are multiple potential matches, both "eth" cropland rows get the first
# "eth" fallowland row, rather than their correct value. The same occurs for the two "sdn" and "bel" rows. The solution is simple:
# joining by iso and countries (or country.codes) instead of simply iso. Alternatively, aggregating to the iso level in each
# FAO table (L100.FAO_X) BEFORE doing further joins and calculations should also lead to the correct answer.
# Take the FAO cropland table, L100.FAO_CL_kha:
L100.FAO_CL_kha %>%
# only include data in the right fallow land year range
filter(year %in% aglu.FALLOW_YEARS) %>%
# keep only the iso country and the value for each:
select(iso, countries, cropland = value, year) %>%
# append in fallow land data in aglu.FALLOW_YEARS from FAO, L100.FAO_fallowland_kha, keeping NA values:
left_join(L100.FAO_fallowland_kha, by = c("iso", "countries", "year")) %>%
# rename value to fallow and remove NAs:
rename(fallow = value) %>%
na.omit() %>%
# add GCAM region information from iso_GCAM_regID
mutate(GCAM_region_ID = left_join(., iso_GCAM_regID, by = "iso")[['GCAM_region_ID']]) %>%
select(GCAM_region_ID, cropland, fallow) %>%
ungroup() %>%
# aggregate cropland and fallow values to the GCAM region level:
group_by(GCAM_region_ID) %>%
summarise_all(sum) %>%
# calculate the fraction of total land in each GCAM region that is fallow:
mutate(fallow_frac = fallow / cropland) ->
# store in a table of cropland, fallow information by region:
L122.cropland_fallow_R
# Lines 55-74 in original file
# Calculating the average percent of cropland that is not in active crop rotations in each region
# make table with cropped land compared to total arable land
# FAO harvested area data, L100.FAO_harv_CL_kha, is appended to the FAO cropland data table. Quantities are aggregated
# from iso to GCAM region and the fraction of cropped land is calculated.
#
# These lines of code also pull FAO data that is susceptible to the double labeling in Ethiopia, Sudan, and Belgium.
# The same match error occurs as above, and we fix in the same way:
# Take the FAO cropland table, L100.FAO_CL_kha:
L100.FAO_CL_kha %>%
# only include data in the right fallow land year range
filter(year %in% aglu.FALLOW_YEARS) %>%
# keep only the iso country and the value for each:
select(iso, countries, cropland = value, year) %>%
# append in cropped land data in aglu.FALLOW_YEARS from FAO, L100.FAO_harv_CL_kha, keeping NA values:
left_join(L100.FAO_harv_CL_kha, by = c("iso", "countries", "year")) %>%
# rename value to cropped and remove NAs:
rename(cropped = value) %>%
na.omit() %>%
# add GCAM region information from iso_GCAM_regID
mutate(GCAM_region_ID = left_join(., iso_GCAM_regID, by = "iso")[['GCAM_region_ID']]) %>%
# remove all columns that are not GCAM_region_ID, cropland, and cropped values:
select(GCAM_region_ID, cropland, cropped) %>%
ungroup() %>%
# aggregate cropland and cropped values to the GCAM region level:
group_by(GCAM_region_ID) %>%
summarise_all(sum) %>%
# calculate the fraction of total land in each GCAM region that is cropped:
mutate(cropped_frac = cropped/cropland) ->
# store in a table of cropland, cropped information by region:
L122.cropland_cropped_R
# Lines 76-89 in original file
# calculate the average amount of cropland that is not in production as the fallow land fraction,
# where available, or else the non-cropped cropland
# based on availability, using (1) fallow land fraction, (2) fraction of cropland not in crop rotations, or (3) 0
# Calculating cropland not in production is done via a series of nested if statements.
# take the unique GCAM regions
tibble::tibble(GCAM_region_ID = unique(iso_GCAM_regID$GCAM_region_ID)) %>%
# sort the ids:
arrange(GCAM_region_ID) %>%
# add a land type identifier for each region:
mutate(Land_Type = "Cropland") %>%
# join in fallow land table, L122.cropland_fallow_R, to get the fallow_frac for each region,
# maintaining NA's for use in determining data availability to calculate nonharvested_frac:
mutate(fallow_frac = left_join(., L122.cropland_fallow_R, by = "GCAM_region_ID")[['fallow_frac']]) %>%
# join in cropped land table, L122.cropland_cropped_R, to get the cropped_frac for each region, and
# use to calculate the uncropped fraction, maintaining NA's to set to 0 later:
mutate(uncropped_frac = 1 - left_join(., L122.cropland_cropped_R, by = "GCAM_region_ID")[['cropped_frac']]) %>%
# calculate the nonharvested fraction of land via nested if statements
# based on availability, using (1) fallow land fraction, (2) land not in crop rotations, or (3) 0
# This step determines (1) versus (2):
# If there is no available fallow land data
mutate(nonharvested_frac = if_else(is.na(fallow_frac),
# use the uncropped fraction
uncropped_frac,
# otherwise, use the available fallow land data:
fallow_frac)) %>%
# This step determines (3) when (1) and (2) are not available:
replace_na(list(nonharvested_frac = 0)) ->
# store in a table of nonharvested cropland fractions by region:
L122.nonharvested_cropland_R
# Lines 91-99 in original file
# applying regional average fallow fractions to all GLUs within each region
# making two tables: the fallow land table and then, using the fallow land table, the available cropland table
# take cropland table with region, landtype and glu information for each year:
L122.LC_bm2_R_CropLand_Y_GLU %>%
# join in the non harvested fraction for each region:
# Note that there is no information for GCAM region 30 in L122.LC_bm2_R_CropLand_Y_GLU, even though
# a nonharvested fraction (of 0) is calculated in L122.nonharvested_cropland_R.
left_join(L122.nonharvested_cropland_R, by = c("GCAM_region_ID", "Land_Type")) %>%
# drop unnecessary columns that come from the join:
select(-fallow_frac, -uncropped_frac) %>%
# use the nonharvested_frac and the cropland area information (value) for each Region-GLU-Year to calculate
# the amouunt of fallowland in that Region-GLU-Year = value * nonharvested_frac
mutate(value = value * nonharvested_frac,
# update the Land_Type identifier to reflect that this is now FallowLand, not Cropland
Land_Type = "FallowLand") %>%
# drop nonharvested_frac since no longer necessary:
select(-nonharvested_frac) ->
# store in a table of fallowland land cover in bm2 for each region-glu-year
L122.LC_bm2_R_FallowLand_Y_GLU
# Use the just made fallow land table to calculate a table of Available CropLand in each region-glu-year
# Available cropland = totalcropland from L122.LC_bm2_R_CropLand_Y_GLU - fallowland from L122.LC_bm2_R_FallowLand_Y_GLU
# Take the table of total cropland in each region-GLU-year:
L122.LC_bm2_R_CropLand_Y_GLU %>%
# join in the land that is fallow in each region-glu-year
# absolutely should match up and it's a problem at this point if they don't, so left_join_error_no_match:
left_join_error_no_match(L122.LC_bm2_R_FallowLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = value from Cropland table, the total amount of cropland in the region-glu-year
# value.y = value from FallowLand table, the amount of fallowland in the region-glu-year
# therefore, available cropland value = value.x-value.y:
mutate(value = value.x-value.y) %>%
# remove value.x, value.y and Land_Type.y as now unnecessary:
select(-value.x, -value.y, -Land_Type.y) %>%
# rename Land_Type.x to Land_Type:
rename(Land_Type = Land_Type.x) ->
# store in a table of available cropland:
L122.LC_bm2_R_AvailableCropLand_Y_GLU
# Lines 101-114 in original file
# Calculate the harvested to cropped land ratio for all crops, by region, year, and GLU
# applying minimum and maximum harvested:cropped ratios
# Maximum harvested:cropped ratio may cause cropland to expand
# This additional cropland will need to be balanced by a deduction from other land types later on, so it is tracked below
# take harvested area by region-glu-year:
L122.ag_HA_bm2_R_Y_GLU %>%
# join the available cropland by region-glu-year
left_join_error_no_match(L122.LC_bm2_R_AvailableCropLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = value from L122.ag_HA_bm2_R_Y_GLU, the harvested area
# value.y = value from L122.LC_bm2_R_AvailableCropLand_Y_GLU, the available cropland
# harvested area to available cropland ratio is value.x/value.y
mutate(value = value.x / value.y) %>%
# remove unneeded columns:
select(-value.x, -value.y, -Land_Type) %>%
# if the harvested to cropland ratio, value, is less than the min acceptable harvested area to cropland,
# aglu.MIN_HA_TO_CROPLAND, replace with aglu.MIN_HA_TO_CROPLAND
mutate(value = if_else(value < aglu.MIN_HA_TO_CROPLAND, aglu.MIN_HA_TO_CROPLAND, value),
# if the harvested to cropland ratio is greater than the max acceptable harvested area to cropland,
# aglu.MAX_HA_TO_CROPLAND, replace with aglu.MAX_HA_TO_CROPLAND
value = if_else(value > aglu.MAX_HA_TO_CROPLAND, aglu.MAX_HA_TO_CROPLAND, value)) ->
# store in a table of HA to cropland ratios by region-glu-year
L122.ag_HA_to_CropLand_R_Y_GLU
# Lines 116-126 in original file
# The ag_HA_to_CropLand ratio is assumed to be a property of the region and GLU (not specific to individual crops)
# Calculate cropland requirements of each crop as harvested area divided by regional ag_HA_to_CropLand ratio
# Take input harvested area by region-commodity-glu-year:
L101.ag_HA_bm2_R_C_Y_GLU %>%
# join the harvested area to cropland ratios, L122.ag_HA_to_CropLand_R_Y_GLU:
left_join_error_no_match(L122.ag_HA_to_CropLand_R_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = original harvested area info for each GCAM_commodity from L101.ag_HA_bm2_R_C_Y_GLU
# value.y = HA:CL ratio from L122.ag_HA_to_CropLand_R_Y_GLU just joined
# calculate the harvested cropland for each commodity as value = value.x/value.y:
mutate(value = value.x / value.y) %>%
select(-value.x, -value.y) %>%
# add a Land_Type identifier:
mutate(Land_Type = "HarvCropLand") ->
# store in a table of Harvested Cropland by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Y_GLU
# Lines 125:128 in original file
# Aggregate across crops to get harvested cropland area, by region/GLU/year
# The harvested cropland area by region-commodity-glu-year table, L122.LC_bm2_R_HarvCropLand_C_Y_GLU,
# is used for other calculations below. That is why this is not part of the previous pipeline
# Take harvested cropland area by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Y_GLU %>%
group_by(GCAM_region_ID, GLU, Land_Type, year) %>%
summarise(value = sum(value)) ->
# store in a table of Harvested cropland by region-glu-year:
L122.LC_bm2_R_HarvCropLand_Y_GLU
# Lines 130-136 in original file
# Calculate economic yield by each crop as production divided by cropland. Write out this preliminary table.
# Production by region-glu-commodity-year comes from input data L101.ag_Prod_Mt_R_C_Y_GLU.
# Cropland area by region-glu-commodity-year was calculated as table L122.LC_bm2_R_HarvCropLand_C_Y_GLU
# Take cropland by region-glu-commodity-year:
L122.LC_bm2_R_HarvCropLand_C_Y_GLU %>%
# join the production by region-glu-commodity-year information:
left_join_error_no_match(L101.ag_Prod_Mt_R_C_Y_GLU, by = c("GCAM_region_ID", "GCAM_commodity", "GCAM_subsector", "GLU", "year")) %>%
# value.x = the cropland area by region-commodity-glu-year from L122.LC_bm2_R_HarvCropLand_C_Y_GLU
# value.y = production by region-commodity-glu-year from L101.ag_Prod_Mt_R_C_Y_GLU
# Calculate yield by region-commodity-glu-year as value = value.y/value.x:
mutate(value = value.y / value.x) %>%
select(-value.x, -value.y, -Land_Type) %>%
# set any NA values to 0 yield:
replace_na(list(value = 0)) ->
# store in a table of Economic Yield by region-commodity-glu-year:
L122.ag_EcYield_kgm2_R_C_Y_GLU
# Calculating OtherArableLand:
# Lines 138:148 in original file
# The minimum threshold on HA:CL means that some cropland in Hyde will not be assigned to a crop. This is mapped to "other arable land"
# The maximum threshold on HA:CL means that cropland in Hyde in some ag regions may be less than cropland in GCAM
# In the latter case, land needs to be mapped to cropland from other uses.
# The first step in executing the three above lines of comment is to calculate the residual land cover that is
# cropland (may be positive or negative): residual = cropland - fallow land - harvested area by crop type
# Take harvested cropland by region-glu-year:
L122.LC_bm2_R_HarvCropLand_Y_GLU %>%
# update the landtype to be residual cropland
ungroup() %>% mutate(Land_Type = "ResidualCropLand") %>%
# Join in the available cropland by region-glu-year
left_join_error_no_match(L122.LC_bm2_R_AvailableCropLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = region-glu-year harvested cropland from L122.LC_bm2_R_HarvCropLand_Y_GLU
# value.y = region-glu-year available cropland from L122.LC_bm2_R_AvailableCropLand_Y_GLU
# Calculate residual cropland as harvested-available, value = value.x-value.y:
mutate(value = value.x - value.y) %>%
# remove unnecessary columns:
select(-value.x, -value.y, -Land_Type.y) %>%
rename(Land_Type = Land_Type.x) ->
# store in a table of Residual Cropland by region-glu-year
L122.LC_bm2_R_ResidualCropLand_Y_GLU
# second step to execute max/min comments above is to process the residuals
# Lines 150-161 in original file
# Calculating unused cropland; this is added with fallow land to calculate other arable land
# Where residuals are negative, this is "unused" cropland that will be mapped to other arable land
# Take the residual cropland by region-glu-year:
L122.LC_bm2_R_ResidualCropLand_Y_GLU %>%
# update the Land_Type identifier
ungroup() %>% mutate(Land_Type = "UnusedCropLand") %>%
# positive residuals cannot be remapped to OtherArableLand. Therefore, they do not count as UnusedCropLand
# and region-glu-years with positive residual have 0 UnusedCropLand:
mutate(value = if_else(value > 0, 0, value),
# negative residuals ARE UnusedCropLand. -1 * the negative residual is the amount of UnusuedCropLand that
# can be mapped to OtherArable Land. Calculate UnusedCropLand:
value = -1 * value) ->
# store in a table of UnusedCropLand by region-glu-year
L122.LC_bm2_R_UnusedCropLand_Y_GLU
# Calculating extra cropland; this will be balanced by a deduction from unmanaged lands
# Where residuals are positive, this is "extra" land that will later be deducted from other categories.
# Take the residual cropland by region-glu-year:
L122.LC_bm2_R_ResidualCropLand_Y_GLU %>%
# update the Land_Type identifier
ungroup() %>% mutate(Land_Type = "ExtraCropLand") %>%
# negative residuals do not count as ExtraCropLand. Therefore, any negative values in value are
# remapped to 0 to represent 0 ExtraCropLand:
mutate(value = if_else(value < 0, 0, value)) ->
# store in a table of ExtraCropLand by region-glu-year:
L122.LC_bm2_R_ExtraCropLand_Y_GLU
# Lines 163-178 in original file
# Calculating other arable land: known fallow plus discrepancy between sum of harvested area and cropland
# Assigning all cropland to other arable land wherever harvested area is zero
# If there are any land use regions with 0 harvested area (Monfreda) but positive cropland cover (Hyde), these are missing
# values in the above table, and all of this cropland should be assigned to other arable land.
#
# Take FallowLand by region-GLU-year:
L122.LC_bm2_R_FallowLand_Y_GLU %>%
# update the Land_Type identifier
ungroup() %>% mutate(Land_Type = "OtherArableLand") %>%
# join the table of unused cropland by region-glu-year, L122.LC_bm2_UnusedCropLand_Y_GLU
# NAs retained as in old DS:
left_join(L122.LC_bm2_R_UnusedCropLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = fallow land in region-glu-year from L122.LC_bm2_R_FallowLand_Y_GLU
# value.y = unused land in region-glu-year from L122.LC_bm2_R_FallowLand_Y_GLU
# OtherArableLand = fallow + unused <=> value = value.x + value.y:
mutate(value = value.x + value.y) %>%
# remove unnecessary columns:
select(-value.x, -value.y, -Land_Type.y) %>%
rename(Land_Type = Land_Type.x) %>%
# join the cropland info so that can handle NA's
left_join(L122.LC_bm2_R_CropLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = OtherArableLand value
# value.y = Cropland value
# Assigning cropland to other arable land wherever harvested area is zero
# If there are any land use regions with 0 harvested area (Monfreda) but positive cropland cover (Hyde), these are missing
# values in the above table, and all of this cropland should be assigned to other arable land.
# for NA values in OtherArableLand (value.x), replace with the cropland value (value.y):
mutate(value = if_else(is.na(value.x), value.y, value.x)) %>%
# remove unnecessary columns
select(-value.x, -value.y, -Land_Type.y) %>%
rename(Land_Type = Land_Type.x) ->
# store in a table of OtherArableLand by region-glu-year:
L122.LC_bm2_R_OtherArableLand_Y_GLU
# Land use history
# Lines 181 - 200 in original file
# Cropland quantities prior to the first AGLU historical year, for spinup of simple carbon cycle model
# NOTE: Simply assigning this to other arable land
# This method differs from prior versions of GCAM, where 1971 land cover quantities were rolled back on the basis of the cropland ratios
# between each land history year and 1971. The problem with this method is that zero values in 1971 can not match non-zero values in prior years.
# The current method instead will have apparent land use change in going from the land history years to the model base years, but due to equal
# soil carbon contents, and similar/small vegetative carbon contents, the net emissions signal should be negligible.
# First, make a table with cropland in the pre-aglu years
L120.LC_bm2_R_LT_Yh_GLU %>%
ungroup %>%
# only save the years in pre-AGLU years:
filter(year %in% aglu.PREAGLU_YEARS) %>% ungroup() %>%
# insure that there is cropland for each GCAM region-glu that appear L122.LC_bm2_R_CropLand_Y_GLU:
tidyr::complete(Land_Type = unique(L122.LC_bm2_R_CropLand_Y_GLU[['Land_Type']]),
tidyr::nesting(GCAM_region_ID, GLU, year), fill = list(value = NA)) %>%
unique() %>%
# join this historical cropland information to the region-glu-landtypes of L122.LC_bm2_R_CropLand_Y_GLU, preserving
# NAs as in old DS:
left_join(
unique(select(L122.LC_bm2_R_CropLand_Y_GLU, GCAM_region_ID, GLU,
Land_Type)),
.,
by = c("GCAM_region_ID", "GLU", "Land_Type")) %>%
# missing values are overwritten to 0:
replace_na(list(value = 0)) %>%
# bind to the table of OtherArableLand information by region-glu-year:
bind_rows(L122.LC_bm2_R_OtherArableLand_Y_GLU, .) %>%
# update the landtype identifier
mutate(Land_Type = "OtherArableLand") ->
# store in a table of OtherArableLand by region-glu-year, including historical years:
L122.LC_bm2_R_OtherArableLand_Yh_GLU
# Lines 202-205 in original file
# All other cropland uses are zero in the pre-aglu years
# ExtraCropLand is expanded to include historical years in each region-glu, with a value of 0.
# Take ExtraCropLand by region-glu-year:
L122.LC_bm2_R_ExtraCropLand_Y_GLU %>%
# expand to include history years and fill in those values to be 0:
tidyr::complete(year = c(aglu.PREAGLU_YEARS, aglu.AGLU_HISTORICAL_YEARS),
nesting(GCAM_region_ID, GLU, Land_Type),
fill = list(value = 0)) ->
# store in a table of ExtraCropland by region-glu-year, including historical years:
L122.LC_bm2_R_ExtraCropLand_Yh_GLU
# Lines 207-210 in original file
# Harvested cropland history
# HarvCropLand is expanded to include historical years for each region-glu-commodity, with a value of 0.
# Take HarvCropLand by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Y_GLU %>%
ungroup %>%
# expand to include history years and fill in those values to be 0:
tidyr::complete(year = c(aglu.PREAGLU_YEARS, aglu.AGLU_HISTORICAL_YEARS),
nesting(GCAM_region_ID, GCAM_commodity, GCAM_subsector, GLU, Land_Type),
fill = list(value = 0)) ->
# store in a table of HarvCropland by region-commodity-glu-year, including historical years:
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU
# Lines 212-217 in original file
# Combine crop types to get land use history by all harvested cropland.
# Use the table of HarvCropland by region-commodity-glu-year, including historical years, L122.LC_bm2_R_HarvCropLand_C_Yh_GLU,
# and aggregate across commodity to arrive at a table of All Harvested Cropland by region-glu-year, including historical years.
# Take HarvCropland by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU %>%
# remove commodity because that is what aggregate over:
select(-GCAM_commodity) %>%
# group by region-glu-year to aggregate:
group_by(GCAM_region_ID, GLU, Land_Type, year) %>%
# aggregate:
summarise(value = sum(value)) ->
# store in a table of all HarvCropland by region-glu-year, including historical years:
L122.LC_bm2_R_HarvCropLand_Yh_GLU
# remove the now-unneeded Land_Type identifier from HarvCropland by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU %>%
select(-Land_Type) ->
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU
# Force-Convert years and GCAM_region_IDs in some problematic outputs to integers:
L122.LC_bm2_R_ExtraCropLand_Yh_GLU %>%
ungroup() %>%
mutate(year = as.integer(year),
GCAM_region_ID = as.integer(GCAM_region_ID)) ->
L122.LC_bm2_R_ExtraCropLand_Yh_GLU
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU %>%
ungroup() %>%
mutate(year = as.integer(year),
GCAM_region_ID = as.integer(GCAM_region_ID)) ->
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU
L122.LC_bm2_R_HarvCropLand_Yh_GLU %>%
ungroup() %>%
mutate(year = as.integer(year),
GCAM_region_ID = as.integer(GCAM_region_ID)) ->
L122.LC_bm2_R_HarvCropLand_Yh_GLU
# Produce outputs
L122.ag_HA_to_CropLand_R_Y_GLU %>%
add_title("Harvested area to cropland ratio by GCAM region / year / GLU") %>%
add_units("Unitless ratio") %>%
add_comments("FAO harvested area, fallow area and land cover area, with Monfreda harvested area and Hyde") %>%
add_comments("land cover data are used to calculate harvested area:cropland for every GCAM region-GLU") %>%
add_legacy_name("L122.ag_HA_to_CropLand_R_Y_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.ag_HA_to_CropLand_R_Y_GLU
L122.ag_EcYield_kgm2_R_C_Y_GLU %>%
add_title("Economic yield by GCAM region / commodity / year / GLU") %>%
add_units("kilogram per meter squared (kg.m2)") %>%
add_comments("Monfreda agricultural production data is combined with aggregated harvested area data at the") %>%
add_comments("region-GLU-commodity level to calculate economic yield.") %>%
add_legacy_name("L122.ag_EcYield_kgm2_R_C_Y_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L101.ag_Prod_Mt_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.ag_EcYield_kgm2_R_C_Y_GLU
L122.LC_bm2_R_OtherArableLand_Yh_GLU %>%
add_title("Other arable land cover by GCAM region / year / GLU") %>%
add_units("billion meters squared (bm2)") %>%
add_comments("Other Arable Land cover is calculated at the region-GLU level by using FAO fallow land data and ") %>%
add_comments("adding the difference Harvested Cropland - Available Cropland. ") %>%
add_legacy_name("L122.LC_bm2_R_OtherArableLand_Yh_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.LC_bm2_R_OtherArableLand_Yh_GLU
L122.LC_bm2_R_ExtraCropLand_Yh_GLU %>%
add_title("Extra cropland cover by GCAM region / year / GLU") %>%
add_units("billion meters squared (bm2)") %>%
add_comments("Any land at the region-GLU level for which Harvested Cropland - Available Cropland is positive, ") %>%
add_comments("making it extra.") %>%
add_legacy_name("L122.LC_bm2_R_ExtraCropLand_Yh_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.LC_bm2_R_ExtraCropLand_Yh_GLU
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU %>%
add_title("Harvested cropland cover by GCAM region / commodity / year / GLU") %>%
add_units("billion meters squared (bm2)") %>%
add_comments("Harvested cropland at the region-GLU-commodity level is calculated by dividing Monfreda harvested area") %>%
add_comments("values by the HarvestedArea:CropLand ratio calculated from Monfreda, FAO, and Hyde land cover and ") %>%
add_comments("harvested area data.") %>%
add_legacy_name("L122.LC_bm2_R_HarvCropLand_C_Yh_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU
L122.LC_bm2_R_HarvCropLand_Yh_GLU %>%
add_title("Harvested cropland cover by GCAM region / year / GLU") %>%
add_units("billion meters squared (bm2)") %>%
add_comments("Harvested cropland at the region-GLU-commodity level is calculated by dividing Monfreda harvested area") %>%
add_comments("values by the HarvestedArea:CropLand ratio calculated from Monfreda, FAO, and Hyde land cover and ") %>%
add_comments("harvested area data. This commodity information is then aggregated to the region-GLU level.") %>%
add_legacy_name("L122.LC_bm2_R_HarvCropLand_Yh_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.LC_bm2_R_HarvCropLand_Yh_GLU
return_data(L122.ag_HA_to_CropLand_R_Y_GLU, L122.ag_EcYield_kgm2_R_C_Y_GLU, L122.LC_bm2_R_OtherArableLand_Yh_GLU, L122.LC_bm2_R_ExtraCropLand_Yh_GLU, L122.LC_bm2_R_HarvCropLand_C_Yh_GLU, L122.LC_bm2_R_HarvCropLand_Yh_GLU)
} else {
stop("Unknown command")
}
}
JGCRI/gcamdata documentation built on March 21, 2023, 2:19 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions module_aglu_LB122.LC_R_Cropland_Yh_GLU
Documented in module_aglu_LB122.LC_R_Cropland_Yh_GLU
# Copyright 2019 Battelle Memorial Institute; see the LICENSE file.
#' module_aglu_LB122.LC_R_Cropland_Yh_GLU
#'
#' Integrate disparate data sources for land cover and harvested area from FAO, Monfreda, and Hyde in
#' order to calculate cropland cover by specific crop type, other arable land, harvested:cropped ratio, economic
#' yield (i.e., production per unit cropland area per year, rather than production per unit area per harvest), and
#' "extra" cropland that needs to be taken from other land use types. All data tables are written out
#' at the GCAM region-GLU level for each historical year, including carbon cycle model spinup years.
#'
#' @param command API command to execute
#' @param ... other optional parameters, depending on command
#' @return Depends on \code{command}: either a vector of required inputs,
#' a vector of output names, or (if \code{command} is "MAKE") all
#' the generated outputs: \code{L122.ag_HA_to_CropLand_R_Y_GLU}, \code{L122.ag_EcYield_kgm2_R_C_Y_GLU},
#' \code{L122.LC_bm2_R_OtherArableLand_Yh_GLU}, \code{L122.LC_bm2_R_ExtraCropLand_Yh_GLU},
#' \code{L122.LC_bm2_R_HarvCropLand_C_Yh_GLU}, \code{L122.LC_bm2_R_HarvCropLand_Yh_GLU}.
#' The corresponding file in the original data system was \code{LB122.LC_R_Cropland_Yh_GLU.R} (aglu level1).
#' @details First, Hyde cropland cover is aggregated to the region-GLU level in each year, and supplemented with Monfreda
#' harvested area information when necessary. Second, FAO fallowland, cropland, and harvested area are aggregated from
#' the country level to the region level, and used to calculate the "residual" cropland in each land use region:
#' total cropland (Hyde) - known fallow (FAO) - harvested crops (FAO/Monfreda).
#' The harvested area : cropland ratio (HA:CL) is computed as (sum of harvested area) / (cropland - fallow land).
#' Where HA:CL < 1, HA:CL is set to 1, and the balance (the "residual" above) is added to OtherArableLand.
#' Where HA:CL > 3, HA:CL is set to 3, and additional cropland (ExtraCropLand) is written out, to be taken from
#' other land use types (determined in different code chunks).
#' @importFrom assertthat assert_that
#' @importFrom dplyr anti_join arrange bind_rows distinct filter if_else group_by left_join mutate select summarise summarise_all
#' @importFrom tidyr replace_na
#' @author ACS April 2017
module_aglu_LB122.LC_R_Cropland_Yh_GLU <- function(command, ...) {
if(command == driver.DECLARE_INPUTS) {
return(c(FILE = "common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L101.ag_Prod_Mt_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU"))
} else if(command == driver.DECLARE_OUTPUTS) {
return(c("L122.ag_HA_to_CropLand_R_Y_GLU",
"L122.ag_EcYield_kgm2_R_C_Y_GLU",
"L122.LC_bm2_R_OtherArableLand_Yh_GLU",
"L122.LC_bm2_R_ExtraCropLand_Yh_GLU",
"L122.LC_bm2_R_HarvCropLand_C_Yh_GLU",
"L122.LC_bm2_R_HarvCropLand_Yh_GLU"))
} else if(command == driver.MAKE) {
Land_Type <- year <- . <- GCAM_region_ID <- GLU <- GCAM_commodity <-
value <- iso <- countries <- cropland <- fallow <- fallow_frac <- cropped <-
cropped_frac <- uncropped_frac <- nonharvested_frac <- value.x <-
value.y <- Land_Type.y <- Land_Type.x <- GCAM_subsector <- NULL # silence package check.
all_data <- list(...)[[1]]
# Load required inputs
iso_GCAM_regID <- get_data(all_data, "common/iso_GCAM_regID")
L100.FAO_fallowland_kha <- get_data(all_data, "L100.FAO_fallowland_kha")
L100.FAO_CL_kha <- get_data(all_data, "L100.FAO_CL_kha")
L100.FAO_harv_CL_kha <- get_data(all_data, "L100.FAO_harv_CL_kha")
L101.ag_HA_bm2_R_C_Y_GLU <- get_data(all_data, "L101.ag_HA_bm2_R_C_Y_GLU", strip_attributes = TRUE)
L101.ag_Prod_Mt_R_C_Y_GLU <- get_data(all_data, "L101.ag_Prod_Mt_R_C_Y_GLU")
L120.LC_bm2_R_LT_Yh_GLU <- get_data(all_data, "L120.LC_bm2_R_LT_Yh_GLU")
# Line 36 in original file
# take a subset of the land cover table L120.LC_bm2_R_LT_YH_GLU: cropland, and only in aglu historical years
L120.LC_bm2_R_LT_Yh_GLU %>%
filter(Land_Type == "Cropland", year %in% aglu.AGLU_HISTORICAL_YEARS) ->
# store in table Land Cover in bm2 by Region, Year and GLU for Cropland.
L122.LC_bm2_R_CropLand_Y_GLU
# Lines 40-45 in original file
# The harvested area/production tables L103.ag.X (from Monfreda) may have R_GLUs not in the cropland table
# L122.LC_bm2_R_CropLand_Y_GLU (from Hyde), and vice versa.
# Fill out the cropland table to include all R_GLUs in Monfreda:
L122.LC_bm2_R_CropLand_Y_GLU %>%
# find the region-GLU combos in the Monfreda harvested area table L101.ag_HA_bm2_R_C_Y_GLU NOT contained
# in the L122.LC_bm2_R_CropLand_Y_GLU land cover table:
anti_join(L101.ag_HA_bm2_R_C_Y_GLU[,c("GCAM_region_ID", "GLU", "year")], .,
by = c("GCAM_region_ID", "GLU", "year")) %>%
# save only the unique combinations:
unique() %>%
# arrange so rows occur in more sensible order:
arrange(GCAM_region_ID, GLU, year) %>%
# add values of 0 everywhere for now:
mutate(value = 0,
# add a Land_Type identifier:
Land_Type = "Cropland") %>%
# add these rows to the original Land Cover table:
bind_rows(L122.LC_bm2_R_CropLand_Y_GLU) ->
# save as the Land C over table:
L122.LC_bm2_R_CropLand_Y_GLU
# Line 48 in original file
# old comment: compile harvested area across all crops
# Take the input historical harvested area table, L101.ag_HA_bm2_R_C_Y_GLU, and sum over GCAM_commodity, so that each
# region-GLU-year combo has a single value.
L101.ag_HA_bm2_R_C_Y_GLU %>%
group_by(GCAM_region_ID, GLU, year) %>%
# and sum by region-GLU-year:
summarise(value = sum(value)) %>%
ungroup() ->
L122.ag_HA_bm2_R_Y_GLU
# Calculating the average percent of cropland that is fallow in each region
# using the average cropland, fallow land, and land in temporary crops from FAO RESOURCESTAT
# The time series is unreliable, so using the average of all available years between 2008 and 2012
# (and applying to all historical years)
# And continue with calculating average crop land, fallow land, land in temporary crops:
# compile the ratio of "temporary crops" to total arable land in each region
# make table with fallow land compared to total arable land
# FAO fallow land data, L100.FAO_fallowland_kha, is appended to the FAO cropland data table. Quantities are aggregated
# from iso to GCAM region and the fraction of fallowland is calculated.
#
# The old data sytem uses match to append FAO fallow land data to FAO cropland data by iso. However, Ethiopia,
# Sudan, and Belgium-Luxembourg have all split into two distinct countries and FAO produces data for each.
# Specifically, there are two rows with iso = "eth" and two rows with iso = "sdn" and two rows with iso = "bel" in some FAO data.
# Because match only pulls the first value when there are multiple potential matches, both "eth" cropland rows get the first
# "eth" fallowland row, rather than their correct value. The same occurs for the two "sdn" and "bel" rows. The solution is simple:
# joining by iso and countries (or country.codes) instead of simply iso. Alternatively, aggregating to the iso level in each
# FAO table (L100.FAO_X) BEFORE doing further joins and calculations should also lead to the correct answer.
# Take the FAO cropland table, L100.FAO_CL_kha:
L100.FAO_CL_kha %>%
# only include data in the right fallow land year range
filter(year %in% aglu.FALLOW_YEARS) %>%
# keep only the iso country and the value for each:
select(iso, countries, cropland = value, year) %>%
# append in fallow land data in aglu.FALLOW_YEARS from FAO, L100.FAO_fallowland_kha, keeping NA values:
left_join(L100.FAO_fallowland_kha, by = c("iso", "countries", "year")) %>%
# rename value to fallow and remove NAs:
rename(fallow = value) %>%
na.omit() %>%
# add GCAM region information from iso_GCAM_regID
mutate(GCAM_region_ID = left_join(., iso_GCAM_regID, by = "iso")[['GCAM_region_ID']]) %>%
select(GCAM_region_ID, cropland, fallow) %>%
ungroup() %>%
# aggregate cropland and fallow values to the GCAM region level:
group_by(GCAM_region_ID) %>%
summarise_all(sum) %>%
# calculate the fraction of total land in each GCAM region that is fallow:
mutate(fallow_frac = fallow / cropland) ->
# store in a table of cropland, fallow information by region:
L122.cropland_fallow_R
# Lines 55-74 in original file
# Calculating the average percent of cropland that is not in active crop rotations in each region
# make table with cropped land compared to total arable land
# FAO harvested area data, L100.FAO_harv_CL_kha, is appended to the FAO cropland data table. Quantities are aggregated
# from iso to GCAM region and the fraction of cropped land is calculated.
#
# These lines of code also pull FAO data that is susceptible to the double labeling in Ethiopia, Sudan, and Belgium.
# The same match error occurs as above, and we fix in the same way:
# Take the FAO cropland table, L100.FAO_CL_kha:
L100.FAO_CL_kha %>%
# only include data in the right fallow land year range
filter(year %in% aglu.FALLOW_YEARS) %>%
# keep only the iso country and the value for each:
select(iso, countries, cropland = value, year) %>%
# append in cropped land data in aglu.FALLOW_YEARS from FAO, L100.FAO_harv_CL_kha, keeping NA values:
left_join(L100.FAO_harv_CL_kha, by = c("iso", "countries", "year")) %>%
# rename value to cropped and remove NAs:
rename(cropped = value) %>%
na.omit() %>%
# add GCAM region information from iso_GCAM_regID
mutate(GCAM_region_ID = left_join(., iso_GCAM_regID, by = "iso")[['GCAM_region_ID']]) %>%
# remove all columns that are not GCAM_region_ID, cropland, and cropped values:
select(GCAM_region_ID, cropland, cropped) %>%
ungroup() %>%
# aggregate cropland and cropped values to the GCAM region level:
group_by(GCAM_region_ID) %>%
summarise_all(sum) %>%
# calculate the fraction of total land in each GCAM region that is cropped:
mutate(cropped_frac = cropped/cropland) ->
# store in a table of cropland, cropped information by region:
L122.cropland_cropped_R
# Lines 76-89 in original file
# calculate the average amount of cropland that is not in production as the fallow land fraction,
# where available, or else the non-cropped cropland
# based on availability, using (1) fallow land fraction, (2) fraction of cropland not in crop rotations, or (3) 0
# Calculating cropland not in production is done via a series of nested if statements.
# take the unique GCAM regions
tibble::tibble(GCAM_region_ID = unique(iso_GCAM_regID$GCAM_region_ID)) %>%
# sort the ids:
arrange(GCAM_region_ID) %>%
# add a land type identifier for each region:
mutate(Land_Type = "Cropland") %>%
# join in fallow land table, L122.cropland_fallow_R, to get the fallow_frac for each region,
# maintaining NA's for use in determining data availability to calculate nonharvested_frac:
mutate(fallow_frac = left_join(., L122.cropland_fallow_R, by = "GCAM_region_ID")[['fallow_frac']]) %>%
# join in cropped land table, L122.cropland_cropped_R, to get the cropped_frac for each region, and
# use to calculate the uncropped fraction, maintaining NA's to set to 0 later:
mutate(uncropped_frac = 1 - left_join(., L122.cropland_cropped_R, by = "GCAM_region_ID")[['cropped_frac']]) %>%
# calculate the nonharvested fraction of land via nested if statements
# based on availability, using (1) fallow land fraction, (2) land not in crop rotations, or (3) 0
# This step determines (1) versus (2):
# If there is no available fallow land data
mutate(nonharvested_frac = if_else(is.na(fallow_frac),
# use the uncropped fraction
uncropped_frac,
# otherwise, use the available fallow land data:
fallow_frac)) %>%
# This step determines (3) when (1) and (2) are not available:
replace_na(list(nonharvested_frac = 0)) ->
# store in a table of nonharvested cropland fractions by region:
L122.nonharvested_cropland_R
# Lines 91-99 in original file
# applying regional average fallow fractions to all GLUs within each region
# making two tables: the fallow land table and then, using the fallow land table, the available cropland table
# take cropland table with region, landtype and glu information for each year:
L122.LC_bm2_R_CropLand_Y_GLU %>%
# join in the non harvested fraction for each region:
# Note that there is no information for GCAM region 30 in L122.LC_bm2_R_CropLand_Y_GLU, even though
# a nonharvested fraction (of 0) is calculated in L122.nonharvested_cropland_R.
left_join(L122.nonharvested_cropland_R, by = c("GCAM_region_ID", "Land_Type")) %>%
# drop unnecessary columns that come from the join:
select(-fallow_frac, -uncropped_frac) %>%
# use the nonharvested_frac and the cropland area information (value) for each Region-GLU-Year to calculate
# the amouunt of fallowland in that Region-GLU-Year = value * nonharvested_frac
mutate(value = value * nonharvested_frac,
# update the Land_Type identifier to reflect that this is now FallowLand, not Cropland
Land_Type = "FallowLand") %>%
# drop nonharvested_frac since no longer necessary:
select(-nonharvested_frac) ->
# store in a table of fallowland land cover in bm2 for each region-glu-year
L122.LC_bm2_R_FallowLand_Y_GLU
# Use the just made fallow land table to calculate a table of Available CropLand in each region-glu-year
# Available cropland = totalcropland from L122.LC_bm2_R_CropLand_Y_GLU - fallowland from L122.LC_bm2_R_FallowLand_Y_GLU
# Take the table of total cropland in each region-GLU-year:
L122.LC_bm2_R_CropLand_Y_GLU %>%
# join in the land that is fallow in each region-glu-year
# absolutely should match up and it's a problem at this point if they don't, so left_join_error_no_match:
left_join_error_no_match(L122.LC_bm2_R_FallowLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = value from Cropland table, the total amount of cropland in the region-glu-year
# value.y = value from FallowLand table, the amount of fallowland in the region-glu-year
# therefore, available cropland value = value.x-value.y:
mutate(value = value.x-value.y) %>%
# remove value.x, value.y and Land_Type.y as now unnecessary:
select(-value.x, -value.y, -Land_Type.y) %>%
# rename Land_Type.x to Land_Type:
rename(Land_Type = Land_Type.x) ->
# store in a table of available cropland:
L122.LC_bm2_R_AvailableCropLand_Y_GLU
# Lines 101-114 in original file
# Calculate the harvested to cropped land ratio for all crops, by region, year, and GLU
# applying minimum and maximum harvested:cropped ratios
# Maximum harvested:cropped ratio may cause cropland to expand
# This additional cropland will need to be balanced by a deduction from other land types later on, so it is tracked below
# take harvested area by region-glu-year:
L122.ag_HA_bm2_R_Y_GLU %>%
# join the available cropland by region-glu-year
left_join_error_no_match(L122.LC_bm2_R_AvailableCropLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = value from L122.ag_HA_bm2_R_Y_GLU, the harvested area
# value.y = value from L122.LC_bm2_R_AvailableCropLand_Y_GLU, the available cropland
# harvested area to available cropland ratio is value.x/value.y
mutate(value = value.x / value.y) %>%
# remove unneeded columns:
select(-value.x, -value.y, -Land_Type) %>%
# if the harvested to cropland ratio, value, is less than the min acceptable harvested area to cropland,
# aglu.MIN_HA_TO_CROPLAND, replace with aglu.MIN_HA_TO_CROPLAND
mutate(value = if_else(value < aglu.MIN_HA_TO_CROPLAND, aglu.MIN_HA_TO_CROPLAND, value),
# if the harvested to cropland ratio is greater than the max acceptable harvested area to cropland,
# aglu.MAX_HA_TO_CROPLAND, replace with aglu.MAX_HA_TO_CROPLAND
value = if_else(value > aglu.MAX_HA_TO_CROPLAND, aglu.MAX_HA_TO_CROPLAND, value)) ->
# store in a table of HA to cropland ratios by region-glu-year
L122.ag_HA_to_CropLand_R_Y_GLU
# Lines 116-126 in original file
# The ag_HA_to_CropLand ratio is assumed to be a property of the region and GLU (not specific to individual crops)
# Calculate cropland requirements of each crop as harvested area divided by regional ag_HA_to_CropLand ratio
# Take input harvested area by region-commodity-glu-year:
L101.ag_HA_bm2_R_C_Y_GLU %>%
# join the harvested area to cropland ratios, L122.ag_HA_to_CropLand_R_Y_GLU:
left_join_error_no_match(L122.ag_HA_to_CropLand_R_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = original harvested area info for each GCAM_commodity from L101.ag_HA_bm2_R_C_Y_GLU
# value.y = HA:CL ratio from L122.ag_HA_to_CropLand_R_Y_GLU just joined
# calculate the harvested cropland for each commodity as value = value.x/value.y:
mutate(value = value.x / value.y) %>%
select(-value.x, -value.y) %>%
# add a Land_Type identifier:
mutate(Land_Type = "HarvCropLand") ->
# store in a table of Harvested Cropland by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Y_GLU
# Lines 125:128 in original file
# Aggregate across crops to get harvested cropland area, by region/GLU/year
# The harvested cropland area by region-commodity-glu-year table, L122.LC_bm2_R_HarvCropLand_C_Y_GLU,
# is used for other calculations below. That is why this is not part of the previous pipeline
# Take harvested cropland area by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Y_GLU %>%
group_by(GCAM_region_ID, GLU, Land_Type, year) %>%
summarise(value = sum(value)) ->
# store in a table of Harvested cropland by region-glu-year:
L122.LC_bm2_R_HarvCropLand_Y_GLU
# Lines 130-136 in original file
# Calculate economic yield by each crop as production divided by cropland. Write out this preliminary table.
# Production by region-glu-commodity-year comes from input data L101.ag_Prod_Mt_R_C_Y_GLU.
# Cropland area by region-glu-commodity-year was calculated as table L122.LC_bm2_R_HarvCropLand_C_Y_GLU
# Take cropland by region-glu-commodity-year:
L122.LC_bm2_R_HarvCropLand_C_Y_GLU %>%
# join the production by region-glu-commodity-year information:
left_join_error_no_match(L101.ag_Prod_Mt_R_C_Y_GLU, by = c("GCAM_region_ID", "GCAM_commodity", "GCAM_subsector", "GLU", "year")) %>%
# value.x = the cropland area by region-commodity-glu-year from L122.LC_bm2_R_HarvCropLand_C_Y_GLU
# value.y = production by region-commodity-glu-year from L101.ag_Prod_Mt_R_C_Y_GLU
# Calculate yield by region-commodity-glu-year as value = value.y/value.x:
mutate(value = value.y / value.x) %>%
select(-value.x, -value.y, -Land_Type) %>%
# set any NA values to 0 yield:
replace_na(list(value = 0)) ->
# store in a table of Economic Yield by region-commodity-glu-year:
L122.ag_EcYield_kgm2_R_C_Y_GLU
# Calculating OtherArableLand:
# Lines 138:148 in original file
# The minimum threshold on HA:CL means that some cropland in Hyde will not be assigned to a crop. This is mapped to "other arable land"
# The maximum threshold on HA:CL means that cropland in Hyde in some ag regions may be less than cropland in GCAM
# In the latter case, land needs to be mapped to cropland from other uses.
# The first step in executing the three above lines of comment is to calculate the residual land cover that is
# cropland (may be positive or negative): residual = cropland - fallow land - harvested area by crop type
# Take harvested cropland by region-glu-year:
L122.LC_bm2_R_HarvCropLand_Y_GLU %>%
# update the landtype to be residual cropland
ungroup() %>% mutate(Land_Type = "ResidualCropLand") %>%
# Join in the available cropland by region-glu-year
left_join_error_no_match(L122.LC_bm2_R_AvailableCropLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = region-glu-year harvested cropland from L122.LC_bm2_R_HarvCropLand_Y_GLU
# value.y = region-glu-year available cropland from L122.LC_bm2_R_AvailableCropLand_Y_GLU
# Calculate residual cropland as harvested-available, value = value.x-value.y:
mutate(value = value.x - value.y) %>%
# remove unnecessary columns:
select(-value.x, -value.y, -Land_Type.y) %>%
rename(Land_Type = Land_Type.x) ->
# store in a table of Residual Cropland by region-glu-year
L122.LC_bm2_R_ResidualCropLand_Y_GLU
# second step to execute max/min comments above is to process the residuals
# Lines 150-161 in original file
# Calculating unused cropland; this is added with fallow land to calculate other arable land
# Where residuals are negative, this is "unused" cropland that will be mapped to other arable land
# Take the residual cropland by region-glu-year:
L122.LC_bm2_R_ResidualCropLand_Y_GLU %>%
# update the Land_Type identifier
ungroup() %>% mutate(Land_Type = "UnusedCropLand") %>%
# positive residuals cannot be remapped to OtherArableLand. Therefore, they do not count as UnusedCropLand
# and region-glu-years with positive residual have 0 UnusedCropLand:
mutate(value = if_else(value > 0, 0, value),
# negative residuals ARE UnusedCropLand. -1 * the negative residual is the amount of UnusuedCropLand that
# can be mapped to OtherArable Land. Calculate UnusedCropLand:
value = -1 * value) ->
# store in a table of UnusedCropLand by region-glu-year
L122.LC_bm2_R_UnusedCropLand_Y_GLU
# Calculating extra cropland; this will be balanced by a deduction from unmanaged lands
# Where residuals are positive, this is "extra" land that will later be deducted from other categories.
# Take the residual cropland by region-glu-year:
L122.LC_bm2_R_ResidualCropLand_Y_GLU %>%
# update the Land_Type identifier
ungroup() %>% mutate(Land_Type = "ExtraCropLand") %>%
# negative residuals do not count as ExtraCropLand. Therefore, any negative values in value are
# remapped to 0 to represent 0 ExtraCropLand:
mutate(value = if_else(value < 0, 0, value)) ->
# store in a table of ExtraCropLand by region-glu-year:
L122.LC_bm2_R_ExtraCropLand_Y_GLU
# Lines 163-178 in original file
# Calculating other arable land: known fallow plus discrepancy between sum of harvested area and cropland
# Assigning all cropland to other arable land wherever harvested area is zero
# If there are any land use regions with 0 harvested area (Monfreda) but positive cropland cover (Hyde), these are missing
# values in the above table, and all of this cropland should be assigned to other arable land.
#
# Take FallowLand by region-GLU-year:
L122.LC_bm2_R_FallowLand_Y_GLU %>%
# update the Land_Type identifier
ungroup() %>% mutate(Land_Type = "OtherArableLand") %>%
# join the table of unused cropland by region-glu-year, L122.LC_bm2_UnusedCropLand_Y_GLU
# NAs retained as in old DS:
left_join(L122.LC_bm2_R_UnusedCropLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = fallow land in region-glu-year from L122.LC_bm2_R_FallowLand_Y_GLU
# value.y = unused land in region-glu-year from L122.LC_bm2_R_FallowLand_Y_GLU
# OtherArableLand = fallow + unused <=> value = value.x + value.y:
mutate(value = value.x + value.y) %>%
# remove unnecessary columns:
select(-value.x, -value.y, -Land_Type.y) %>%
rename(Land_Type = Land_Type.x) %>%
# join the cropland info so that can handle NA's
left_join(L122.LC_bm2_R_CropLand_Y_GLU, by = c("GCAM_region_ID", "GLU", "year")) %>%
# value.x = OtherArableLand value
# value.y = Cropland value
# Assigning cropland to other arable land wherever harvested area is zero
# If there are any land use regions with 0 harvested area (Monfreda) but positive cropland cover (Hyde), these are missing
# values in the above table, and all of this cropland should be assigned to other arable land.
# for NA values in OtherArableLand (value.x), replace with the cropland value (value.y):
mutate(value = if_else(is.na(value.x), value.y, value.x)) %>%
# remove unnecessary columns
select(-value.x, -value.y, -Land_Type.y) %>%
rename(Land_Type = Land_Type.x) ->
# store in a table of OtherArableLand by region-glu-year:
L122.LC_bm2_R_OtherArableLand_Y_GLU
# Land use history
# Lines 181 - 200 in original file
# Cropland quantities prior to the first AGLU historical year, for spinup of simple carbon cycle model
# NOTE: Simply assigning this to other arable land
# This method differs from prior versions of GCAM, where 1971 land cover quantities were rolled back on the basis of the cropland ratios
# between each land history year and 1971. The problem with this method is that zero values in 1971 can not match non-zero values in prior years.
# The current method instead will have apparent land use change in going from the land history years to the model base years, but due to equal
# soil carbon contents, and similar/small vegetative carbon contents, the net emissions signal should be negligible.
# First, make a table with cropland in the pre-aglu years
L120.LC_bm2_R_LT_Yh_GLU %>%
ungroup %>%
# only save the years in pre-AGLU years:
filter(year %in% aglu.PREAGLU_YEARS) %>% ungroup() %>%
# insure that there is cropland for each GCAM region-glu that appear L122.LC_bm2_R_CropLand_Y_GLU:
tidyr::complete(Land_Type = unique(L122.LC_bm2_R_CropLand_Y_GLU[['Land_Type']]),
tidyr::nesting(GCAM_region_ID, GLU, year), fill = list(value = NA)) %>%
unique() %>%
# join this historical cropland information to the region-glu-landtypes of L122.LC_bm2_R_CropLand_Y_GLU, preserving
# NAs as in old DS:
left_join(
unique(select(L122.LC_bm2_R_CropLand_Y_GLU, GCAM_region_ID, GLU,
Land_Type)),
.,
by = c("GCAM_region_ID", "GLU", "Land_Type")) %>%
# missing values are overwritten to 0:
replace_na(list(value = 0)) %>%
# bind to the table of OtherArableLand information by region-glu-year:
bind_rows(L122.LC_bm2_R_OtherArableLand_Y_GLU, .) %>%
# update the landtype identifier
mutate(Land_Type = "OtherArableLand") ->
# store in a table of OtherArableLand by region-glu-year, including historical years:
L122.LC_bm2_R_OtherArableLand_Yh_GLU
# Lines 202-205 in original file
# All other cropland uses are zero in the pre-aglu years
# ExtraCropLand is expanded to include historical years in each region-glu, with a value of 0.
# Take ExtraCropLand by region-glu-year:
L122.LC_bm2_R_ExtraCropLand_Y_GLU %>%
# expand to include history years and fill in those values to be 0:
tidyr::complete(year = c(aglu.PREAGLU_YEARS, aglu.AGLU_HISTORICAL_YEARS),
nesting(GCAM_region_ID, GLU, Land_Type),
fill = list(value = 0)) ->
# store in a table of ExtraCropland by region-glu-year, including historical years:
L122.LC_bm2_R_ExtraCropLand_Yh_GLU
# Lines 207-210 in original file
# Harvested cropland history
# HarvCropLand is expanded to include historical years for each region-glu-commodity, with a value of 0.
# Take HarvCropLand by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Y_GLU %>%
ungroup %>%
# expand to include history years and fill in those values to be 0:
tidyr::complete(year = c(aglu.PREAGLU_YEARS, aglu.AGLU_HISTORICAL_YEARS),
nesting(GCAM_region_ID, GCAM_commodity, GCAM_subsector, GLU, Land_Type),
fill = list(value = 0)) ->
# store in a table of HarvCropland by region-commodity-glu-year, including historical years:
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU
# Lines 212-217 in original file
# Combine crop types to get land use history by all harvested cropland.
# Use the table of HarvCropland by region-commodity-glu-year, including historical years, L122.LC_bm2_R_HarvCropLand_C_Yh_GLU,
# and aggregate across commodity to arrive at a table of All Harvested Cropland by region-glu-year, including historical years.
# Take HarvCropland by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU %>%
# remove commodity because that is what aggregate over:
select(-GCAM_commodity) %>%
# group by region-glu-year to aggregate:
group_by(GCAM_region_ID, GLU, Land_Type, year) %>%
# aggregate:
summarise(value = sum(value)) ->
# store in a table of all HarvCropland by region-glu-year, including historical years:
L122.LC_bm2_R_HarvCropLand_Yh_GLU
# remove the now-unneeded Land_Type identifier from HarvCropland by region-commodity-glu-year:
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU %>%
select(-Land_Type) ->
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU
# Force-Convert years and GCAM_region_IDs in some problematic outputs to integers:
L122.LC_bm2_R_ExtraCropLand_Yh_GLU %>%
ungroup() %>%
mutate(year = as.integer(year),
GCAM_region_ID = as.integer(GCAM_region_ID)) ->
L122.LC_bm2_R_ExtraCropLand_Yh_GLU
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU %>%
ungroup() %>%
mutate(year = as.integer(year),
GCAM_region_ID = as.integer(GCAM_region_ID)) ->
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU
L122.LC_bm2_R_HarvCropLand_Yh_GLU %>%
ungroup() %>%
mutate(year = as.integer(year),
GCAM_region_ID = as.integer(GCAM_region_ID)) ->
L122.LC_bm2_R_HarvCropLand_Yh_GLU
# Produce outputs
L122.ag_HA_to_CropLand_R_Y_GLU %>%
add_title("Harvested area to cropland ratio by GCAM region / year / GLU") %>%
add_units("Unitless ratio") %>%
add_comments("FAO harvested area, fallow area and land cover area, with Monfreda harvested area and Hyde") %>%
add_comments("land cover data are used to calculate harvested area:cropland for every GCAM region-GLU") %>%
add_legacy_name("L122.ag_HA_to_CropLand_R_Y_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.ag_HA_to_CropLand_R_Y_GLU
L122.ag_EcYield_kgm2_R_C_Y_GLU %>%
add_title("Economic yield by GCAM region / commodity / year / GLU") %>%
add_units("kilogram per meter squared (kg.m2)") %>%
add_comments("Monfreda agricultural production data is combined with aggregated harvested area data at the") %>%
add_comments("region-GLU-commodity level to calculate economic yield.") %>%
add_legacy_name("L122.ag_EcYield_kgm2_R_C_Y_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L101.ag_Prod_Mt_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.ag_EcYield_kgm2_R_C_Y_GLU
L122.LC_bm2_R_OtherArableLand_Yh_GLU %>%
add_title("Other arable land cover by GCAM region / year / GLU") %>%
add_units("billion meters squared (bm2)") %>%
add_comments("Other Arable Land cover is calculated at the region-GLU level by using FAO fallow land data and ") %>%
add_comments("adding the difference Harvested Cropland - Available Cropland. ") %>%
add_legacy_name("L122.LC_bm2_R_OtherArableLand_Yh_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.LC_bm2_R_OtherArableLand_Yh_GLU
L122.LC_bm2_R_ExtraCropLand_Yh_GLU %>%
add_title("Extra cropland cover by GCAM region / year / GLU") %>%
add_units("billion meters squared (bm2)") %>%
add_comments("Any land at the region-GLU level for which Harvested Cropland - Available Cropland is positive, ") %>%
add_comments("making it extra.") %>%
add_legacy_name("L122.LC_bm2_R_ExtraCropLand_Yh_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.LC_bm2_R_ExtraCropLand_Yh_GLU
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU %>%
add_title("Harvested cropland cover by GCAM region / commodity / year / GLU") %>%
add_units("billion meters squared (bm2)") %>%
add_comments("Harvested cropland at the region-GLU-commodity level is calculated by dividing Monfreda harvested area") %>%
add_comments("values by the HarvestedArea:CropLand ratio calculated from Monfreda, FAO, and Hyde land cover and ") %>%
add_comments("harvested area data.") %>%
add_legacy_name("L122.LC_bm2_R_HarvCropLand_C_Yh_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.LC_bm2_R_HarvCropLand_C_Yh_GLU
L122.LC_bm2_R_HarvCropLand_Yh_GLU %>%
add_title("Harvested cropland cover by GCAM region / year / GLU") %>%
add_units("billion meters squared (bm2)") %>%
add_comments("Harvested cropland at the region-GLU-commodity level is calculated by dividing Monfreda harvested area") %>%
add_comments("values by the HarvestedArea:CropLand ratio calculated from Monfreda, FAO, and Hyde land cover and ") %>%
add_comments("harvested area data. This commodity information is then aggregated to the region-GLU level.") %>%
add_legacy_name("L122.LC_bm2_R_HarvCropLand_Yh_GLU") %>%
add_precursors("common/iso_GCAM_regID",
"L100.FAO_fallowland_kha",
"L100.FAO_CL_kha",
"L100.FAO_harv_CL_kha",
"L101.ag_HA_bm2_R_C_Y_GLU",
"L120.LC_bm2_R_LT_Yh_GLU") ->
L122.LC_bm2_R_HarvCropLand_Yh_GLU
return_data(L122.ag_HA_to_CropLand_R_Y_GLU, L122.ag_EcYield_kgm2_R_C_Y_GLU, L122.LC_bm2_R_OtherArableLand_Yh_GLU, L122.LC_bm2_R_ExtraCropLand_Yh_GLU, L122.LC_bm2_R_HarvCropLand_C_Yh_GLU, L122.LC_bm2_R_HarvCropLand_Yh_GLU)
} else {
stop("Unknown command")
}
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.