R/zchunk_L170.nonghg_ceds_scaling_USA.R
In JGCRI/gcamdata: Build the GCAM Input Datasets
Defines functions
module_gcamusa_L170.nonghg_ceds_scaling_USA
Documented in
module_gcamusa_L170.nonghg_ceds_scaling_USA
#' module_gcamusa_L170.nonghg_ceds_scaling_USA
#'
#' Scales pre-processed NEI non-CO2 emissions to CEDS at the national level
#' The NEI data is processed exogenously. It is scaled to EPA Tier 1, interpolated
#' between NEI years, extrapolated back to 1990, and mapped to GCAM sectors,
#' CEDS sectors, and CEDS fuels
#'
#'
#' @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{L170.NEI_1990_2017_GCAM_sectors}, \code{L170.NEI_CEDS_scaling_diagnostic}
#' @details Non-CO2 emissions from NEI scaled to CEDS
#' @importFrom assertthat assert_that
#' @importFrom dplyr filter mutate select group_by
#' @importFrom tidyr gather spread
#' @author MAW February 2021
module_gcamusa_L170.nonghg_ceds_scaling_USA <- function(command, ...) {
if(command == driver.DECLARE_INPUTS) {
return(c(FILE="gcam-usa/emissions/NEI_pollutant_mapping",
FILE="gcam-usa/emissions/CEDS_to_CEDS_sector_mapping",
"L169.NEI_1990_2017_GCAM_sectors_unscaled",
"L102.ceds_GFED_nonco2_tg_C_S_F"))
} else if(command == driver.DECLARE_OUTPUTS) {
return(c("L170.NEI_1990_2017_GCAM_sectors",
"L170.NEI_CEDS_scaling_diagnostic"))
} else if(command == driver.MAKE) {
# Silence package check.
X <- NULL
all_data <- list(...)[[1]]
# Load required inputs
NEI_pollutant_mapping <- get_data(all_data, "gcam-usa/emissions/NEI_pollutant_mapping")
CEDS_to_CEDS_sector_mapping <- get_data(all_data, "gcam-usa/emissions/CEDS_to_CEDS_sector_mapping")
L102.ceds_GFED_nonco2_tg_C_S_F <- get_data(all_data, "L102.ceds_GFED_nonco2_tg_C_S_F")
L169.NEI_1990_2017_GCAM_sectors_unscaled <- get_data(all_data, "L169.NEI_1990_2017_GCAM_sectors_unscaled", strip_attributes = T)
# -----------------------------------------------------------------------------
# Perform computations
# Getting mapping file in order
# Remove Notes column
CEDS_to_CEDS <- CEDS_to_CEDS_sector_mapping %>%
select( -Notes )
# Remove Core_CEDS column from NEI mapping and NA entries so that we can use left_join_error_no_match
US_SCC_CEDS_to_Agg <- CEDS_to_CEDS %>%
select( -Core_CEDS ) %>%
filter( !is.na( SCC_CEDS_extended ) )
# Some Core_CEDS sectors have duplicate entries due to being further disaggregated in SCC_CEDS_extended sectors
# ex. Core_CEDS = 1A1bc_Other-transformation maps to three SCC_CEDS_extended sectors
# to remediate this, we need to remove the SCC_CEDS_extended column for the CEDS_to_CEDS mapping
Core_CEDS_to_Agg <- CEDS_to_CEDS %>%
select( -SCC_CEDS_extended ) %>%
distinct( Core_CEDS, Agg_CEDS_for_scaling )
# Convert CEDS from Tg to US Ton
ceds_TON <- L102.ceds_GFED_nonco2_tg_C_S_F %>%
mutate( emissions = emissions / CONV_TST_TG * 1000,
units = "TON" )
# ========== ROAD
# The Road sector has a discontinuity in emissions, so we use a different scaling approach for BC, OC, PM2.5
# From 2008 forward, we will scale CEDS to NEI
# From 2007 back to 1990, we will scale CEDS using the 2008 scaling factor and linearly interpolate it to
# Create a table that contains all of the NEI PM2.5 Road emissions
NEI_road <- L169.NEI_1990_2017_GCAM_sectors_unscaled %>%
left_join_error_no_match( US_SCC_CEDS_to_Agg, by = c( "CEDS_Sector" = "SCC_CEDS_extended" ) ) %>%
left_join_error_no_match( NEI_pollutant_mapping, by = c( "pollutant" = "NEI_pollutant" ) ) %>%
filter( Agg_CEDS_for_scaling == "1A3b_Road",
Non.CO2 == "PM2.5",
!is.na( CEDS_Fuel) )
# 2008 forward
# Sum NEI emissions by year, pollutant, and CEDS_Fuel
NEI_PM2.5_road_sums_2008 <- NEI_road %>%
filter( Agg_CEDS_for_scaling == "1A3b_Road",
Non.CO2 == "PM2.5",
year > 2007 ) %>%
group_by( year, Non.CO2, CEDS_Fuel ) %>%
mutate( NEI_sum = sum(emissions) ) %>%
distinct( year, Non.CO2, NEI_sum, CEDS_Fuel )
# Sum CEDS emissions by year, pollutant, and fuel
CEDS_PM2.5_road_sums_2008 <- ceds_TON %>%
filter( sector == "1A3b_Road",
Non.CO2 == "BC" | Non.CO2 == "OC",
year > 2007 ) %>%
group_by( year, Non.CO2, fuel ) %>%
mutate( CEDS_sum = sum(emissions) ) %>%
distinct( year, Non.CO2, CEDS_sum, fuel ) %>%
# make wide to calculate PM2.5
spread( Non.CO2, CEDS_sum ) %>%
# because CEDS does not have PM2.5, convert sum of BC + OC to PM2.5
mutate( "PM2.5" = ( BC + OC * gcamusa.OC_TO_OM ) * gcamusa.PM1_TO_PM2.5 ) %>%
# make long again
tidyr::gather( Non.CO2, CEDS_sum, -c( year, fuel ) ) %>%
filter( Non.CO2 == "PM2.5" )
# Join the tables to get scaling factors to apply to PM2.5, OC, and BC
road_scaling <- NEI_PM2.5_road_sums_2008 %>%
left_join( CEDS_PM2.5_road_sums_2008, by = c( "year", "Non.CO2", "CEDS_Fuel" = "fuel" ) ) %>%
mutate( scaling_factor = CEDS_sum / NEI_sum ) %>%
ungroup() %>%
select( year, CEDS_Fuel, scaling_factor )
# Scale CEDS BC and OC using this factor
CEDS_road_scaled_2008 <- ceds_TON %>%
filter( iso == "usa",
sector == "1A3b_Road",
Non.CO2 == "BC" | Non.CO2 == "OC",
year > 2007 ) %>%
left_join( road_scaling, by = c( "year", "fuel" = "CEDS_Fuel" ) ) %>%
mutate( emissions = emissions / scaling_factor ) %>%
# filter out NAs (these are years there isn't NEI data for, and also fuels with 0 emissions that become NA)
filter( !is.na( emissions ) ) %>%
select( -scaling_factor )
# 2007 back
# use the 2008 scaling factor, and linearly interpolate it to a constant in 1990
road_scaling_1990_BC <- road_scaling %>%
filter( year == 2008 ) %>%
mutate( year = 1990,
Non.CO2 = "BC",
scaling_factor = gcamusa.BC_1990_ONROAD_SCALING_FACTOR )
road_scaling_1990_BC_OC <- road_scaling_1990_BC %>%
mutate( Non.CO2 = "OC",
scaling_factor = gcamusa.OC_1990_ONROAD_SCALING_FACTOR ) %>%
bind_rows( road_scaling_1990_BC)
road_scaling_2008 <- road_scaling %>%
filter( year == 2008 ) %>%
repeat_add_columns(tibble::tibble(Non.CO2=unique(road_scaling_1990_BC_OC$Non.CO2)))
road_scaling_2007_back <- road_scaling_2008 %>%
bind_rows( road_scaling_1990_BC_OC ) %>%
complete( nesting( CEDS_Fuel, Non.CO2 ), year = min(road_scaling_1990_BC_OC$year):max(road_scaling_2008$year)) %>%
group_by( CEDS_Fuel, Non.CO2 ) %>%
mutate( scaling_factor = approx_fun(year, scaling_factor, rule = 2) ) %>%
unique()
CEDS_road_scaled_2007_back <- ceds_TON %>%
filter( iso == "usa",
sector == "1A3b_Road",
Non.CO2 == "BC" | Non.CO2 == "OC",
year < 2008 ) %>%
left_join( road_scaling_2007_back, by = c( "year", "fuel" = "CEDS_Fuel", "Non.CO2" ) ) %>%
mutate( emissions = emissions / scaling_factor ) %>%
select( -scaling_factor ) %>%
# several NAs for fuels that have 0 emissions (brown_coal, coal_coke, etc.)
# filter out NAs
filter( !is.na( emissions ) )
# Bind CEDS tables back together
BC_OC_Road_CEDS <- CEDS_road_scaled_2007_back %>%
bind_rows( CEDS_road_scaled_2008 ) %>%
filter( emissions != 0 )
# Table with NEI Road PM 2.5 for 2007 and before, to be scaled to CEDS nationally
NEI_PM2.5_road_2007 <- NEI_road %>%
filter( Agg_CEDS_for_scaling == "1A3b_Road",
Non.CO2 == "PM2.5",
year < 2008 )
# Calculating PM2.5 equivalents for 2007 and before using CEDS BC and OC
CEDS_PM2.5_road_sums_2007 <- BC_OC_Road_CEDS %>%
filter( iso == "usa",
sector == "1A3b_Road",
Non.CO2 == "BC" | Non.CO2 == "OC",
year < 2008 ) %>%
# remove NA emissions in order for "sum" to work
filter( !is.na( emissions ) ) %>%
group_by( year, Non.CO2, fuel ) %>%
mutate( CEDS_sum = sum(emissions) ) %>%
distinct( year, Non.CO2, CEDS_sum, fuel ) %>%
# make wide to calculate PM2.5
spread( Non.CO2, CEDS_sum ) %>%
mutate( "PM2.5" = ( BC + OC * gcamusa.OC_TO_OM ) * gcamusa.PM1_TO_PM2.5 ) %>%
# make long again
tidyr::gather( Non.CO2, CEDS_sum, -c( year, fuel ) ) %>%
filter( Non.CO2 == "PM2.5" )
# Join the CEDS sums with the NEI PM2.5 values
NEI_PM2.5_road_scaled_2007 <- NEI_PM2.5_road_2007 %>%
# group emissions by year and fuel to get NEI_sum
group_by( year, CEDS_Fuel ) %>%
mutate( NEI_sum = sum( emissions ) ) %>%
left_join( CEDS_PM2.5_road_sums_2007, by = c( "year", "Non.CO2", "CEDS_Fuel" = "fuel" ) ) %>%
mutate( scaling_factor = NEI_sum / CEDS_sum,
emissions = emissions/ scaling_factor ) %>%
select( -c( NEI_sum, CEDS_sum, scaling_factor ) ) %>%
# natural gas is all NAs since not in CEDS
filter( !is.na(emissions) )
# BIND AND FORMAT
# Bind NEI PM2.5 2008 and later with NEI scaled PM2.5 2007 and before
NEI_PM2.5_road <- NEI_road %>%
filter( Agg_CEDS_for_scaling == "1A3b_Road",
Non.CO2 == "PM2.5",
year > 2007 ) %>%
bind_rows( NEI_PM2.5_road_scaled_2007 ) %>%
select( -pollutant )
# Get CEDS BC and OC at state level
# take CEDS national BC OC PM2.5 ratio and apply that to the states by CEDS fuel (diesel vs gasoline)
fuel_ratio <- BC_OC_Road_CEDS %>%
group_by( year, Non.CO2, fuel ) %>%
mutate( CEDS_sum = sum(emissions) ) %>%
# select relevant columns
select( c( year, Non.CO2, CEDS_sum, fuel ) ) %>%
# make wide to calculate PM2.5
spread( Non.CO2, CEDS_sum ) %>%
mutate( "PM2.5" = ( BC + OC * gcamusa.OC_TO_OM ) * gcamusa.PM1_TO_PM2.5,
# calculate BC and OC shares as a fraction of PM2.5
"BC_frac" = BC / PM2.5,
"OC_frac" = OC / PM2.5 ) %>%
# select relevant columns
select( -c( BC, OC, PM2.5 ) )
# apply the fractions to the state level emissions data, basing BC and OC emissions
# off of PM2.5 and the ratios calculated from CEDS
road <- NEI_PM2.5_road %>%
# rename "emissions" column PM2.5, as these are PM2.5 emissions
rename( PM2.5 = emissions ) %>%
left_join( fuel_ratio, by = c( "year", "CEDS_Fuel" = "fuel" ) ) %>%
# calculate BC and OC emissions at the state level using the ratios
mutate( BC = PM2.5 * BC_frac,
OC = PM2.5 * OC_frac ) %>%
# deselct the ratio columns, and gather the pollutant columns
select( -c( BC_frac, OC_frac ) ) %>%
tidyr::gather( key = "Non.CO2", value = "emissions", "PM2.5", "BC", "OC" ) %>%
filter( !is.na( emissions ) )
# ========== NON-ROAD
# Biodiesel and Ethanol Production do not have emissions in CEDS, so these
# sectors will be filtered out and we will use emissions straight from NEI
# Petroleum Refining is in CEDS, but we will use emissions straight from NEI
# because the scaling reduced these emissions too much
biodiesel_ethanol_prod_petr_refining <- L169.NEI_1990_2017_GCAM_sectors_unscaled %>%
left_join_error_no_match( US_SCC_CEDS_to_Agg, by = c( "CEDS_Sector" = "SCC_CEDS_extended" ) ) %>%
left_join_error_no_match( NEI_pollutant_mapping, by = c( "pollutant" = "NEI_pollutant" ) ) %>%
select( -c( pollutant ) ) %>%
filter( Agg_CEDS_for_scaling == "1A1bc_Other-transformation" |
CEDS_Sector == "1A1b_Pet-refining")
# For all other sectors / pollutants other than Road PM2.5, we will scale to CEDS
# First, bind mapping files to NEI table so this only has to be done once
NEI_all_unscaled_no_road <- L169.NEI_1990_2017_GCAM_sectors_unscaled %>%
left_join_error_no_match( US_SCC_CEDS_to_Agg, by = c( "CEDS_Sector" = "SCC_CEDS_extended" ) ) %>%
left_join_error_no_match( NEI_pollutant_mapping, by = c( "pollutant" = "NEI_pollutant" ) ) %>%
select( -c( pollutant ) ) %>%
# filter out PM2.5 from the road sectors
filter( Non.CO2 != "PM2.5" | Agg_CEDS_for_scaling != "1A3b_Road",
Agg_CEDS_for_scaling != "1A1bc_Other-transformation",
CEDS_Sector != "1A1b_Pet-refining" )
# Create a table that has CEDS sums by year, aggregate sector, and Non.CO2
# This will be compared to NEI sums by year, aggregate sector, and Non.CO2
CEDS_sums <- ceds_TON %>%
filter( iso == "usa",
# filter out Road BC and OC
sector != "1A3b_Road" | Non.CO2 != "BC" & Non.CO2 != "OC" ) %>%
left_join_error_no_match( Core_CEDS_to_Agg, by = c( "sector" = "Core_CEDS") ) %>%
select( -c( sector, CEDS_agg_sector, CEDS_agg_fuel ) ) %>%
# group by year and pollutant, and sum emissions
dplyr::group_by( year, Non.CO2, Agg_CEDS_for_scaling ) %>%
dplyr::mutate( CEDS_emissions = sum( emissions ) ) %>%
distinct( year, CEDS_emissions, Non.CO2, Agg_CEDS_for_scaling )
# Create a table that has NEI sums by year, aggregate sector, and Non.CO2
NEI_sums <- NEI_all_unscaled_no_road %>%
dplyr::group_by( year, Non.CO2, Agg_CEDS_for_scaling ) %>%
dplyr::mutate( NEI_emissions = sum( emissions ) ) %>%
distinct( year, NEI_emissions, Non.CO2, Agg_CEDS_for_scaling ) %>%
select( c( "year", "NEI_emissions", "Agg_CEDS_for_scaling", "Non.CO2" ) )
# calculate an NEI to CEDS scaling factor for each year to multiply NEI emissions by
NEI_CEDS_scaling <- NEI_sums %>%
# can't use left_join_error_no_match because NEI has pollutants CEDS doesn't and vice versa
# CEDS has BC, CH4, OC, N20, H2
# NEI has PM2.5, PM10
# pollutants in both: NH3, CO, NOx, SO2, NMVOC
left_join( CEDS_sums, by = c( "year", "Non.CO2", "Agg_CEDS_for_scaling" ) ) %>%
filter( Agg_CEDS_for_scaling != "not in gcam",
Non.CO2 == "NH3" | Non.CO2 == "CO" | Non.CO2 == "NOx" | Non.CO2 == "SO2" | Non.CO2 == "NMVOC" ) %>%
mutate( scaling_factor = NEI_emissions/CEDS_emissions )
# scaling factors are "Inf" when NEI has emissions that are 0 in CEDS, NaN when emissions are 0 in both,
# and 0 when CEDS has emissions not in NEI
# diagnostic table to check scaling factors
NEI_CEDS_scaling_diagnostic <- NEI_CEDS_scaling %>%
select( -c( NEI_emissions, CEDS_emissions ) ) %>%
filter( Non.CO2 != "PM2.5" & Non.CO2 != "PM10" ) %>%
spread( key = year, value = scaling_factor ) %>%
ungroup()
# CEDS doesn't have PM2.5 or PM10, so these pollutants will be filtered out of NEI and not scaled
NEI_noPM <- NEI_all_unscaled_no_road %>%
filter( Non.CO2 != "PM2.5" & Non.CO2 != "PM10",
Agg_CEDS_for_scaling != "not in gcam" ) %>%
left_join( NEI_CEDS_scaling, by = c( "year", "Non.CO2", "Agg_CEDS_for_scaling" ) ) %>%
mutate( emissions = emissions / scaling_factor ) %>%
select( -c( NEI_emissions, CEDS_emissions, scaling_factor ) )
# Table containing PM2.5 and PM10 emissions, bound with other emissions that were scaled to CEDS
NEI_1990_2017_GCAM_sectors <- NEI_all_unscaled_no_road %>%
filter( Non.CO2 == "PM2.5" | Non.CO2 == "PM10" ) %>%
bind_rows( NEI_noPM ) %>%
mutate( unit = "TON" ) %>%
# add back the Road, biodiesel and ethanol production, and petroleum refining emissions
bind_rows( road, biodiesel_ethanol_prod_petr_refining ) %>%
# 5C_Open-burning-land-clearing are not used in GCAM, and not mapped to a GCAM Sector
filter( !is.na( GCAM_sector ), GCAM_sector != 0 )
# NA entries (emissions) become 0 in order for emissions sums to work (otherwise NA is returned)
# The NAs are in Agg_CEDS_for_scaling "3B_Manure-management", for CO, NOx, and SO2 in 2014 - 2017
# where the scaling factor is NaN or 0
NEI_1990_2017_GCAM_sectors <- as.data.frame(NEI_1990_2017_GCAM_sectors)
NEI_1990_2017_GCAM_sectors[is.na(NEI_1990_2017_GCAM_sectors)] <- 0
NEI_1990_2017_GCAM_sectors <- as_tibble(NEI_1990_2017_GCAM_sectors)
# Process into gcam datasystem format
NEI_1990_2017_GCAM_sectors_final <- NEI_1990_2017_GCAM_sectors %>%
left_join_error_no_match( NEI_pollutant_mapping, by = c( "Non.CO2" ) ) %>%
rename( "pollutant" = "NEI_pollutant" ) %>%
select( -c( Non.CO2, Agg_CEDS_for_scaling, CEDS_Sector ) )
# ===================================================
# Produce outputs
L170.NEI_1990_2017_GCAM_sectors <- NEI_1990_2017_GCAM_sectors_final %>%
add_title("Non-CO2 emissions by US state / GCAM sector / fuel / pollutant / year") %>%
add_units("TON") %>%
add_comments("Psuedo-NEI emissions that have been scaled to CEDS at the national level, after being scaled to state level tier1 data in an exogenous script") %>%
add_precursors("gcam-usa/emissions/NEI_pollutant_mapping",
"gcam-usa/emissions/CEDS_to_CEDS_sector_mapping",
"L102.ceds_GFED_nonco2_tg_C_S_F",
"L169.NEI_1990_2017_GCAM_sectors_unscaled")
L170.NEI_CEDS_scaling_diagnostic <- NEI_CEDS_scaling_diagnostic %>%
add_title("Scaling factors used to scale NEI emissions to match CEDS nationally (USA) by aggregate sector") %>%
add_units("Unitless") %>%
add_comments("These factors can be used in post processing, particularly for 1A3aii_Domestic-aviation and 1A3dii_Domestic-navigation (shipping)") %>%
add_comments("These sectors include different things between NEI and CEDS, so can be used to convert back to what is in NEI. Factor is for NEI/CEDS") %>%
add_precursors("gcam-usa/emissions/NEI_pollutant_mapping",
"gcam-usa/emissions/CEDS_to_CEDS_sector_mapping",
"L102.ceds_GFED_nonco2_tg_C_S_F",
"L169.NEI_1990_2017_GCAM_sectors_unscaled")
return_data(L170.NEI_1990_2017_GCAM_sectors, L170.NEI_CEDS_scaling_diagnostic)
} 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_gcamusa_L170.nonghg_ceds_scaling_USA
Documented in module_gcamusa_L170.nonghg_ceds_scaling_USA
#' module_gcamusa_L170.nonghg_ceds_scaling_USA
#'
#' Scales pre-processed NEI non-CO2 emissions to CEDS at the national level
#' The NEI data is processed exogenously. It is scaled to EPA Tier 1, interpolated
#' between NEI years, extrapolated back to 1990, and mapped to GCAM sectors,
#' CEDS sectors, and CEDS fuels
#'
#'
#' @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{L170.NEI_1990_2017_GCAM_sectors}, \code{L170.NEI_CEDS_scaling_diagnostic}
#' @details Non-CO2 emissions from NEI scaled to CEDS
#' @importFrom assertthat assert_that
#' @importFrom dplyr filter mutate select group_by
#' @importFrom tidyr gather spread
#' @author MAW February 2021
module_gcamusa_L170.nonghg_ceds_scaling_USA <- function(command, ...) {
if(command == driver.DECLARE_INPUTS) {
return(c(FILE="gcam-usa/emissions/NEI_pollutant_mapping",
FILE="gcam-usa/emissions/CEDS_to_CEDS_sector_mapping",
"L169.NEI_1990_2017_GCAM_sectors_unscaled",
"L102.ceds_GFED_nonco2_tg_C_S_F"))
} else if(command == driver.DECLARE_OUTPUTS) {
return(c("L170.NEI_1990_2017_GCAM_sectors",
"L170.NEI_CEDS_scaling_diagnostic"))
} else if(command == driver.MAKE) {
# Silence package check.
X <- NULL
all_data <- list(...)[[1]]
# Load required inputs
NEI_pollutant_mapping <- get_data(all_data, "gcam-usa/emissions/NEI_pollutant_mapping")
CEDS_to_CEDS_sector_mapping <- get_data(all_data, "gcam-usa/emissions/CEDS_to_CEDS_sector_mapping")
L102.ceds_GFED_nonco2_tg_C_S_F <- get_data(all_data, "L102.ceds_GFED_nonco2_tg_C_S_F")
L169.NEI_1990_2017_GCAM_sectors_unscaled <- get_data(all_data, "L169.NEI_1990_2017_GCAM_sectors_unscaled", strip_attributes = T)
# -----------------------------------------------------------------------------
# Perform computations
# Getting mapping file in order
# Remove Notes column
CEDS_to_CEDS <- CEDS_to_CEDS_sector_mapping %>%
select( -Notes )
# Remove Core_CEDS column from NEI mapping and NA entries so that we can use left_join_error_no_match
US_SCC_CEDS_to_Agg <- CEDS_to_CEDS %>%
select( -Core_CEDS ) %>%
filter( !is.na( SCC_CEDS_extended ) )
# Some Core_CEDS sectors have duplicate entries due to being further disaggregated in SCC_CEDS_extended sectors
# ex. Core_CEDS = 1A1bc_Other-transformation maps to three SCC_CEDS_extended sectors
# to remediate this, we need to remove the SCC_CEDS_extended column for the CEDS_to_CEDS mapping
Core_CEDS_to_Agg <- CEDS_to_CEDS %>%
select( -SCC_CEDS_extended ) %>%
distinct( Core_CEDS, Agg_CEDS_for_scaling )
# Convert CEDS from Tg to US Ton
ceds_TON <- L102.ceds_GFED_nonco2_tg_C_S_F %>%
mutate( emissions = emissions / CONV_TST_TG * 1000,
units = "TON" )
# ========== ROAD
# The Road sector has a discontinuity in emissions, so we use a different scaling approach for BC, OC, PM2.5
# From 2008 forward, we will scale CEDS to NEI
# From 2007 back to 1990, we will scale CEDS using the 2008 scaling factor and linearly interpolate it to
# Create a table that contains all of the NEI PM2.5 Road emissions
NEI_road <- L169.NEI_1990_2017_GCAM_sectors_unscaled %>%
left_join_error_no_match( US_SCC_CEDS_to_Agg, by = c( "CEDS_Sector" = "SCC_CEDS_extended" ) ) %>%
left_join_error_no_match( NEI_pollutant_mapping, by = c( "pollutant" = "NEI_pollutant" ) ) %>%
filter( Agg_CEDS_for_scaling == "1A3b_Road",
Non.CO2 == "PM2.5",
!is.na( CEDS_Fuel) )
# 2008 forward
# Sum NEI emissions by year, pollutant, and CEDS_Fuel
NEI_PM2.5_road_sums_2008 <- NEI_road %>%
filter( Agg_CEDS_for_scaling == "1A3b_Road",
Non.CO2 == "PM2.5",
year > 2007 ) %>%
group_by( year, Non.CO2, CEDS_Fuel ) %>%
mutate( NEI_sum = sum(emissions) ) %>%
distinct( year, Non.CO2, NEI_sum, CEDS_Fuel )
# Sum CEDS emissions by year, pollutant, and fuel
CEDS_PM2.5_road_sums_2008 <- ceds_TON %>%
filter( sector == "1A3b_Road",
Non.CO2 == "BC" | Non.CO2 == "OC",
year > 2007 ) %>%
group_by( year, Non.CO2, fuel ) %>%
mutate( CEDS_sum = sum(emissions) ) %>%
distinct( year, Non.CO2, CEDS_sum, fuel ) %>%
# make wide to calculate PM2.5
spread( Non.CO2, CEDS_sum ) %>%
# because CEDS does not have PM2.5, convert sum of BC + OC to PM2.5
mutate( "PM2.5" = ( BC + OC * gcamusa.OC_TO_OM ) * gcamusa.PM1_TO_PM2.5 ) %>%
# make long again
tidyr::gather( Non.CO2, CEDS_sum, -c( year, fuel ) ) %>%
filter( Non.CO2 == "PM2.5" )
# Join the tables to get scaling factors to apply to PM2.5, OC, and BC
road_scaling <- NEI_PM2.5_road_sums_2008 %>%
left_join( CEDS_PM2.5_road_sums_2008, by = c( "year", "Non.CO2", "CEDS_Fuel" = "fuel" ) ) %>%
mutate( scaling_factor = CEDS_sum / NEI_sum ) %>%
ungroup() %>%
select( year, CEDS_Fuel, scaling_factor )
# Scale CEDS BC and OC using this factor
CEDS_road_scaled_2008 <- ceds_TON %>%
filter( iso == "usa",
sector == "1A3b_Road",
Non.CO2 == "BC" | Non.CO2 == "OC",
year > 2007 ) %>%
left_join( road_scaling, by = c( "year", "fuel" = "CEDS_Fuel" ) ) %>%
mutate( emissions = emissions / scaling_factor ) %>%
# filter out NAs (these are years there isn't NEI data for, and also fuels with 0 emissions that become NA)
filter( !is.na( emissions ) ) %>%
select( -scaling_factor )
# 2007 back
# use the 2008 scaling factor, and linearly interpolate it to a constant in 1990
road_scaling_1990_BC <- road_scaling %>%
filter( year == 2008 ) %>%
mutate( year = 1990,
Non.CO2 = "BC",
scaling_factor = gcamusa.BC_1990_ONROAD_SCALING_FACTOR )
road_scaling_1990_BC_OC <- road_scaling_1990_BC %>%
mutate( Non.CO2 = "OC",
scaling_factor = gcamusa.OC_1990_ONROAD_SCALING_FACTOR ) %>%
bind_rows( road_scaling_1990_BC)
road_scaling_2008 <- road_scaling %>%
filter( year == 2008 ) %>%
repeat_add_columns(tibble::tibble(Non.CO2=unique(road_scaling_1990_BC_OC$Non.CO2)))
road_scaling_2007_back <- road_scaling_2008 %>%
bind_rows( road_scaling_1990_BC_OC ) %>%
complete( nesting( CEDS_Fuel, Non.CO2 ), year = min(road_scaling_1990_BC_OC$year):max(road_scaling_2008$year)) %>%
group_by( CEDS_Fuel, Non.CO2 ) %>%
mutate( scaling_factor = approx_fun(year, scaling_factor, rule = 2) ) %>%
unique()
CEDS_road_scaled_2007_back <- ceds_TON %>%
filter( iso == "usa",
sector == "1A3b_Road",
Non.CO2 == "BC" | Non.CO2 == "OC",
year < 2008 ) %>%
left_join( road_scaling_2007_back, by = c( "year", "fuel" = "CEDS_Fuel", "Non.CO2" ) ) %>%
mutate( emissions = emissions / scaling_factor ) %>%
select( -scaling_factor ) %>%
# several NAs for fuels that have 0 emissions (brown_coal, coal_coke, etc.)
# filter out NAs
filter( !is.na( emissions ) )
# Bind CEDS tables back together
BC_OC_Road_CEDS <- CEDS_road_scaled_2007_back %>%
bind_rows( CEDS_road_scaled_2008 ) %>%
filter( emissions != 0 )
# Table with NEI Road PM 2.5 for 2007 and before, to be scaled to CEDS nationally
NEI_PM2.5_road_2007 <- NEI_road %>%
filter( Agg_CEDS_for_scaling == "1A3b_Road",
Non.CO2 == "PM2.5",
year < 2008 )
# Calculating PM2.5 equivalents for 2007 and before using CEDS BC and OC
CEDS_PM2.5_road_sums_2007 <- BC_OC_Road_CEDS %>%
filter( iso == "usa",
sector == "1A3b_Road",
Non.CO2 == "BC" | Non.CO2 == "OC",
year < 2008 ) %>%
# remove NA emissions in order for "sum" to work
filter( !is.na( emissions ) ) %>%
group_by( year, Non.CO2, fuel ) %>%
mutate( CEDS_sum = sum(emissions) ) %>%
distinct( year, Non.CO2, CEDS_sum, fuel ) %>%
# make wide to calculate PM2.5
spread( Non.CO2, CEDS_sum ) %>%
mutate( "PM2.5" = ( BC + OC * gcamusa.OC_TO_OM ) * gcamusa.PM1_TO_PM2.5 ) %>%
# make long again
tidyr::gather( Non.CO2, CEDS_sum, -c( year, fuel ) ) %>%
filter( Non.CO2 == "PM2.5" )
# Join the CEDS sums with the NEI PM2.5 values
NEI_PM2.5_road_scaled_2007 <- NEI_PM2.5_road_2007 %>%
# group emissions by year and fuel to get NEI_sum
group_by( year, CEDS_Fuel ) %>%
mutate( NEI_sum = sum( emissions ) ) %>%
left_join( CEDS_PM2.5_road_sums_2007, by = c( "year", "Non.CO2", "CEDS_Fuel" = "fuel" ) ) %>%
mutate( scaling_factor = NEI_sum / CEDS_sum,
emissions = emissions/ scaling_factor ) %>%
select( -c( NEI_sum, CEDS_sum, scaling_factor ) ) %>%
# natural gas is all NAs since not in CEDS
filter( !is.na(emissions) )
# BIND AND FORMAT
# Bind NEI PM2.5 2008 and later with NEI scaled PM2.5 2007 and before
NEI_PM2.5_road <- NEI_road %>%
filter( Agg_CEDS_for_scaling == "1A3b_Road",
Non.CO2 == "PM2.5",
year > 2007 ) %>%
bind_rows( NEI_PM2.5_road_scaled_2007 ) %>%
select( -pollutant )
# Get CEDS BC and OC at state level
# take CEDS national BC OC PM2.5 ratio and apply that to the states by CEDS fuel (diesel vs gasoline)
fuel_ratio <- BC_OC_Road_CEDS %>%
group_by( year, Non.CO2, fuel ) %>%
mutate( CEDS_sum = sum(emissions) ) %>%
# select relevant columns
select( c( year, Non.CO2, CEDS_sum, fuel ) ) %>%
# make wide to calculate PM2.5
spread( Non.CO2, CEDS_sum ) %>%
mutate( "PM2.5" = ( BC + OC * gcamusa.OC_TO_OM ) * gcamusa.PM1_TO_PM2.5,
# calculate BC and OC shares as a fraction of PM2.5
"BC_frac" = BC / PM2.5,
"OC_frac" = OC / PM2.5 ) %>%
# select relevant columns
select( -c( BC, OC, PM2.5 ) )
# apply the fractions to the state level emissions data, basing BC and OC emissions
# off of PM2.5 and the ratios calculated from CEDS
road <- NEI_PM2.5_road %>%
# rename "emissions" column PM2.5, as these are PM2.5 emissions
rename( PM2.5 = emissions ) %>%
left_join( fuel_ratio, by = c( "year", "CEDS_Fuel" = "fuel" ) ) %>%
# calculate BC and OC emissions at the state level using the ratios
mutate( BC = PM2.5 * BC_frac,
OC = PM2.5 * OC_frac ) %>%
# deselct the ratio columns, and gather the pollutant columns
select( -c( BC_frac, OC_frac ) ) %>%
tidyr::gather( key = "Non.CO2", value = "emissions", "PM2.5", "BC", "OC" ) %>%
filter( !is.na( emissions ) )
# ========== NON-ROAD
# Biodiesel and Ethanol Production do not have emissions in CEDS, so these
# sectors will be filtered out and we will use emissions straight from NEI
# Petroleum Refining is in CEDS, but we will use emissions straight from NEI
# because the scaling reduced these emissions too much
biodiesel_ethanol_prod_petr_refining <- L169.NEI_1990_2017_GCAM_sectors_unscaled %>%
left_join_error_no_match( US_SCC_CEDS_to_Agg, by = c( "CEDS_Sector" = "SCC_CEDS_extended" ) ) %>%
left_join_error_no_match( NEI_pollutant_mapping, by = c( "pollutant" = "NEI_pollutant" ) ) %>%
select( -c( pollutant ) ) %>%
filter( Agg_CEDS_for_scaling == "1A1bc_Other-transformation" |
CEDS_Sector == "1A1b_Pet-refining")
# For all other sectors / pollutants other than Road PM2.5, we will scale to CEDS
# First, bind mapping files to NEI table so this only has to be done once
NEI_all_unscaled_no_road <- L169.NEI_1990_2017_GCAM_sectors_unscaled %>%
left_join_error_no_match( US_SCC_CEDS_to_Agg, by = c( "CEDS_Sector" = "SCC_CEDS_extended" ) ) %>%
left_join_error_no_match( NEI_pollutant_mapping, by = c( "pollutant" = "NEI_pollutant" ) ) %>%
select( -c( pollutant ) ) %>%
# filter out PM2.5 from the road sectors
filter( Non.CO2 != "PM2.5" | Agg_CEDS_for_scaling != "1A3b_Road",
Agg_CEDS_for_scaling != "1A1bc_Other-transformation",
CEDS_Sector != "1A1b_Pet-refining" )
# Create a table that has CEDS sums by year, aggregate sector, and Non.CO2
# This will be compared to NEI sums by year, aggregate sector, and Non.CO2
CEDS_sums <- ceds_TON %>%
filter( iso == "usa",
# filter out Road BC and OC
sector != "1A3b_Road" | Non.CO2 != "BC" & Non.CO2 != "OC" ) %>%
left_join_error_no_match( Core_CEDS_to_Agg, by = c( "sector" = "Core_CEDS") ) %>%
select( -c( sector, CEDS_agg_sector, CEDS_agg_fuel ) ) %>%
# group by year and pollutant, and sum emissions
dplyr::group_by( year, Non.CO2, Agg_CEDS_for_scaling ) %>%
dplyr::mutate( CEDS_emissions = sum( emissions ) ) %>%
distinct( year, CEDS_emissions, Non.CO2, Agg_CEDS_for_scaling )
# Create a table that has NEI sums by year, aggregate sector, and Non.CO2
NEI_sums <- NEI_all_unscaled_no_road %>%
dplyr::group_by( year, Non.CO2, Agg_CEDS_for_scaling ) %>%
dplyr::mutate( NEI_emissions = sum( emissions ) ) %>%
distinct( year, NEI_emissions, Non.CO2, Agg_CEDS_for_scaling ) %>%
select( c( "year", "NEI_emissions", "Agg_CEDS_for_scaling", "Non.CO2" ) )
# calculate an NEI to CEDS scaling factor for each year to multiply NEI emissions by
NEI_CEDS_scaling <- NEI_sums %>%
# can't use left_join_error_no_match because NEI has pollutants CEDS doesn't and vice versa
# CEDS has BC, CH4, OC, N20, H2
# NEI has PM2.5, PM10
# pollutants in both: NH3, CO, NOx, SO2, NMVOC
left_join( CEDS_sums, by = c( "year", "Non.CO2", "Agg_CEDS_for_scaling" ) ) %>%
filter( Agg_CEDS_for_scaling != "not in gcam",
Non.CO2 == "NH3" | Non.CO2 == "CO" | Non.CO2 == "NOx" | Non.CO2 == "SO2" | Non.CO2 == "NMVOC" ) %>%
mutate( scaling_factor = NEI_emissions/CEDS_emissions )
# scaling factors are "Inf" when NEI has emissions that are 0 in CEDS, NaN when emissions are 0 in both,
# and 0 when CEDS has emissions not in NEI
# diagnostic table to check scaling factors
NEI_CEDS_scaling_diagnostic <- NEI_CEDS_scaling %>%
select( -c( NEI_emissions, CEDS_emissions ) ) %>%
filter( Non.CO2 != "PM2.5" & Non.CO2 != "PM10" ) %>%
spread( key = year, value = scaling_factor ) %>%
ungroup()
# CEDS doesn't have PM2.5 or PM10, so these pollutants will be filtered out of NEI and not scaled
NEI_noPM <- NEI_all_unscaled_no_road %>%
filter( Non.CO2 != "PM2.5" & Non.CO2 != "PM10",
Agg_CEDS_for_scaling != "not in gcam" ) %>%
left_join( NEI_CEDS_scaling, by = c( "year", "Non.CO2", "Agg_CEDS_for_scaling" ) ) %>%
mutate( emissions = emissions / scaling_factor ) %>%
select( -c( NEI_emissions, CEDS_emissions, scaling_factor ) )
# Table containing PM2.5 and PM10 emissions, bound with other emissions that were scaled to CEDS
NEI_1990_2017_GCAM_sectors <- NEI_all_unscaled_no_road %>%
filter( Non.CO2 == "PM2.5" | Non.CO2 == "PM10" ) %>%
bind_rows( NEI_noPM ) %>%
mutate( unit = "TON" ) %>%
# add back the Road, biodiesel and ethanol production, and petroleum refining emissions
bind_rows( road, biodiesel_ethanol_prod_petr_refining ) %>%
# 5C_Open-burning-land-clearing are not used in GCAM, and not mapped to a GCAM Sector
filter( !is.na( GCAM_sector ), GCAM_sector != 0 )
# NA entries (emissions) become 0 in order for emissions sums to work (otherwise NA is returned)
# The NAs are in Agg_CEDS_for_scaling "3B_Manure-management", for CO, NOx, and SO2 in 2014 - 2017
# where the scaling factor is NaN or 0
NEI_1990_2017_GCAM_sectors <- as.data.frame(NEI_1990_2017_GCAM_sectors)
NEI_1990_2017_GCAM_sectors[is.na(NEI_1990_2017_GCAM_sectors)] <- 0
NEI_1990_2017_GCAM_sectors <- as_tibble(NEI_1990_2017_GCAM_sectors)
# Process into gcam datasystem format
NEI_1990_2017_GCAM_sectors_final <- NEI_1990_2017_GCAM_sectors %>%
left_join_error_no_match( NEI_pollutant_mapping, by = c( "Non.CO2" ) ) %>%
rename( "pollutant" = "NEI_pollutant" ) %>%
select( -c( Non.CO2, Agg_CEDS_for_scaling, CEDS_Sector ) )
# ===================================================
# Produce outputs
L170.NEI_1990_2017_GCAM_sectors <- NEI_1990_2017_GCAM_sectors_final %>%
add_title("Non-CO2 emissions by US state / GCAM sector / fuel / pollutant / year") %>%
add_units("TON") %>%
add_comments("Psuedo-NEI emissions that have been scaled to CEDS at the national level, after being scaled to state level tier1 data in an exogenous script") %>%
add_precursors("gcam-usa/emissions/NEI_pollutant_mapping",
"gcam-usa/emissions/CEDS_to_CEDS_sector_mapping",
"L102.ceds_GFED_nonco2_tg_C_S_F",
"L169.NEI_1990_2017_GCAM_sectors_unscaled")
L170.NEI_CEDS_scaling_diagnostic <- NEI_CEDS_scaling_diagnostic %>%
add_title("Scaling factors used to scale NEI emissions to match CEDS nationally (USA) by aggregate sector") %>%
add_units("Unitless") %>%
add_comments("These factors can be used in post processing, particularly for 1A3aii_Domestic-aviation and 1A3dii_Domestic-navigation (shipping)") %>%
add_comments("These sectors include different things between NEI and CEDS, so can be used to convert back to what is in NEI. Factor is for NEI/CEDS") %>%
add_precursors("gcam-usa/emissions/NEI_pollutant_mapping",
"gcam-usa/emissions/CEDS_to_CEDS_sector_mapping",
"L102.ceds_GFED_nonco2_tg_C_S_F",
"L169.NEI_1990_2017_GCAM_sectors_unscaled")
return_data(L170.NEI_1990_2017_GCAM_sectors, L170.NEI_CEDS_scaling_diagnostic)
} 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.