.Rhistory (169-231-83-144.wireless.ucsb.edu's conflicted copy 2019-12-12)

# Export merged data
outfile_basename <- paste(toupper(rcp), type, "optimal", sep="_")
writeRaster(ras_farm_spp, file=file.path(outputdir, paste0(outfile_basename, "_species.tif")), overwrite=T)
writeRaster(ras_farm_prod, file=file.path(outputdir, paste0(outfile_basename, "_production_mt_yr.tif")), overwrite=T)
writeRaster(ras_farm_profits, file=file.path(outputdir, paste0(outfile_basename, "_profits_usd_yr.tif")), overwrite=T)
}
# Run function
calc_optimal_aqprod(rcp="rcp85", species=species, years=c(2021, 2051, 2100), type="Bivalve")
# Functional programming approach
################################################################################
# # Function to read stack into R
# read_stack <- function(rcp, species){
#   filename <- paste(toupper(rcp), "Bivalve", gsub(" ", "_", species), "profits_usd_yr.tif", sep="_")
#   profit_ras <- raster::brick(file.path(inputdir,  filename))
#   return(profit_ras)
# }
#
# # Build key
# stack_key <- tibble(rcp="rcp85",
#                     species=species) %>%
#   mutate(stack=purrr::map2(rcp, species, read_stack)) %>%
#   mutate(yr2021=list(magrittr::extract2(stack, 1) %>% magrittr::extract2(1)),
#          yr2051=list(magrittr::extract2(stack, 1) %>% magrittr::extract2(2)),
#          yr2100=list(magrittr::extract2(stack, 1) %>% magrittr::extract2(3)))
#
# # I used this code to figure out how to write the function above
# # stack_key$stack[[1]][[1]]
# # stack_key$yr2100
#
# # Build year stacks
# stack_2021 <- stack(stack_key$yr2021)
# stack_2051 <- stack(stack_key$yr2051)
# stack_2100 <- stack(stack_key$yr2100)
#
#
# # Determine which layer (species) is the most profitable
# stack_2021_wmax <- which.max(stack_2021)
# stack_2051_wmax <- which.max(stack_2051)
# stack_2100_wmax <- which.max(stack_2100)
# Clear workspace
rm(list = ls())
# Setup
################################################################################
# Packages
library(raster)
library(ggplot2)
library(tidyverse)
# Directories
outputdir <- "output/processed"
list.files(outputdir)
# Read data
prod <- raster(file.path(outputdir, "RCP85_Bivalve_optimal_production_mt_yr.tif"))
profits <- raster(file.path(outputdir, "RCP85_Bivalve_optimal_profits_usd_yr.tif"))
species <- raster(file.path(outputdir, "RCP85_Bivalve_optimal_species.tif"))
# Clear workspace
rm(list = ls())
# Setup
################################################################################
# Packages
library(raster)
library(ggplot2)
library(tidyverse)
# Directories
outputdir <- "output/processed"
# Read data
prod <- raster(file.path(outputdir, "RCP85_Bivalve_optimal_production_mt_yr.tif"))
profits <- raster(file.path(outputdir, "RCP85_Bivalve_optimal_profits_usd_yr.tif"))
species <- raster(file.path(outputdir, "RCP85_Bivalve_optimal_species.tif"))
plot(species)
plot(profits)
prod <- brick(file.path(outputdir, "RCP85_Bivalve_optimal_production_mt_yr.tif"))
profits <- brick(file.path(outputdir, "RCP85_Bivalve_optimal_profits_usd_yr.tif"))
species <- brick(file.path(outputdir, "RCP85_Bivalve_optimal_species.tif"))
plot(species)
# Plot data
plot(prod)
# Plot data
plot(prod/1e6)
# Clear workspace
rm(list = ls())
# Setup
################################################################################
# Packages
library(raster)
library(ggplot2)
library(tidyverse)
# Directories
codedir <- "code"
sppdir <- "data/species/data"
outputdir <- "output/raw"
# Read aquacast function
source(file.path(codedir, "aquacast.R"))
source(file.path(codedir, "calc_costs.R"))
# Read species data
load(file.path(sppdir, "aquaculture_species_key.Rdata"))
# Setup for testing
################################################################################
# Check error in following species:
# Crassostrea rhizophorae, Crassostrea tulipa, Crassostrea virginica, Ostrea chilensis:  Error in value[j, ] : incorrect number of dimensions
# Clear workspace
rm(list = ls())
# Setup
################################################################################
# Packages
library(raster)
library(ggplot2)
library(tidyverse)
# Directories
codedir <- "code"
sppdir <- "data/species/data"
outputdir <- "output/raw"
# Read aquacast function
source(file.path(codedir, "aquacast.R"))
source(file.path(codedir, "calc_costs.R"))
# Read species data
load(file.path(sppdir, "aquaculture_species_key.Rdata"))
# Setup for testing
################################################################################
# Check error in following species:
# Crassostrea rhizophorae, Crassostrea tulipa, Crassostrea virginica, Ostrea chilensis:  Error in value[j, ] : incorrect number of dimensions
View(data)
# Subset finfish
finfish <- filter(data, class=="Actinopterygii") %>%
mutate(type="Finfish")
1:nrow(finfish)
# Loop through species and run model
i <- 1
# Parameters
species <- finfish[i,]
# years <- 2021
years <- c(2021, 2051, 2100)
rcp="rcp85"
# Read EEZs
eezdir <- "data/eezs"
eezs <- raster(file.path(eezdir, "eezs_v10_raster_10km.tif"))
eez_key <- read.csv(file.path(eezdir, "eezs_v10_key.csv"), as.is=T)
# Build EEZ mask
eez_mask <- eezs
eez_mask[!is.na(eez_mask)] <- 1
# Read climate forecasts
climatedir <- "data/climate/data/gfdl/GFDL-ESM2G/4rasters_scaled"
ras_sst_c_min <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_tos_degC_annual_min_scaled.tif")))
ras_sst_c_max <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_tos_degC_annual_max_scaled.tif")))
ras_sal_psu_mean <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_so_psu_annual_mean_scaled.tif")))
ras_o2_molm3_mean <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_o2_mol_m3_annual_mean_scaled.tif")))
ras_chl_mgm3_meansubsd <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_chl_mg_m3_annual_mean_minus_sd_scaled.tif")))
ras_arag_sat_mean <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_arag_sat_annual_mean_scaled.tif")))
ras_layer_names <- readRDS(file.path(climatedir, "GFDL_ESM2M_rcp85_arag_sat_annual_mean_scaled_layer_names.rds"))
# Add raster layer names
names(ras_sst_c_min) <- names(ras_sst_c_max) <- names(ras_sal_psu_mean) <- names(ras_o2_molm3_mean) <-
names(ras_chl_mgm3_meansubsd) <- names(ras_arag_sat_mean) <- ras_layer_names
# Check rasters
env_ras_check <- compareRaster(eezs, ras_sst_c_min, ras_sst_c_max, ras_sal_psu_mean, ras_o2_molm3_mean, ras_chl_mgm3_meansubsd, ras_arag_sat_mean)
if(env_ras_check==F){print("EEZ and climate forecast rasters DO NOT have the same projection, extent, and resolution.")}
# 1. Extract parameters
####################################
# Growth and harvest parameters
spp <- species$species
type <- species$type
linf_cm <- species$linf_cm
k <- species$k
harvest_cm <- species$harvest_cm
harvest_g <- species$harvest_g
harvest_yr <- species$harvest_yr
a <- species$a
b <- species$b
fcr <- species$fcr
price_usd_mt <- species$price_usd_mt_isscaap
print(spp)
sst_c_min <- species$sst_c_min
sst_c_max <- species$sst_c_max
sal_psu_min <- species$sal_psu_min
sal_psu_max <- species$sal_psu_max
if(type=="Finfish"){
o2_molm3_min <- 0.2757
chl_mgm3_meansubsd_min <- NA
arag_sat_min <- NA
}
if(type=="Bivalve"){
o2_molm3_min <- 0.1244
chl_mgm3_meansubsd_min <- 0.4
arag_sat_min <- 1
}
if(type=="Finfish"){
# Finfish farm design
farm_design <- tibble(type="finfish",
area_sqkm=1,
ncages=24,
cage_vol_m3=9000,
juv_m3=20) %>%
mutate(nstocked=ncages * cage_vol_m3 * juv_m3)
}
# Bivalve farm design
if(type=="Bivalve"){
farm_design <- tibble(type="bivalve",
area_sqkm=1,
nlines=100,
line_m=4000,
juv_ft=100) %>%
mutate(line_ft=measurements::conv_unit(line_m, "m", "ft"),
nstocked=nlines * line_ft * juv_ft)
}
# Calculate yield per year (kg/yr) for 1 sqkm farm
farm_kg_yr <- harvest_g/1000 * farm_design$nstocked / harvest_yr
farm_mt_yr <- farm_kg_yr / 1000
yrs_available <- as.numeric(ras_layer_names)
yrs_do_indices <- which(yrs_available %in% years)
# Viable cells
print("... identifying viable cells")
sst_c_mask <- ras_sst_c_min[[yrs_do_indices]] >= sst_c_min & ras_sst_c_max[[yrs_do_indices]] <= sst_c_max
sal_psu_mask <- ras_sal_psu_mean[[yrs_do_indices]] >= sal_psu_min & ras_sal_psu_mean[[yrs_do_indices]] <= sal_psu_max
# Bivalve
if(type=="Bivalve"){
o2_molm3_mask <- ras_o2_molm3_mean[[yrs_do_indices]] >= o2_molm3_min
chl_mgm3_meansubsd_mask <- ras_chl_mgm3_meansubsd[[yrs_do_indices]] >= chl_mgm3_meansubsd_min
arag_sat_mask <- ras_arag_sat_mean[[yrs_do_indices]] >= arag_sat_min
vcells <- eez_mask * sst_c_mask * sal_psu_mask * o2_molm3_mask * chl_mgm3_meansubsd_mask * arag_sat_mask
NAvalue(vcells) <- 0
}
# Finfish
if(type=="Finfish"){
o2_molm3_mask <- o2_molm3_mean[[yrs_do_indices]] >= o2_molm3_min
vcells <- eez_mask * sst_c_mask * sal_psu_mask * o2_molm3_mask
NAvalue(vcells) <- 0
}
o2_molm3_mask <- ras_o2_molm3_mean[[yrs_do_indices]] >= o2_molm3_min
vcells <- eez_mask * sst_c_mask * sal_psu_mask * o2_molm3_mask
NAvalue(vcells) <- 0
plot(vcells[[1]], main="Suitable cells")
plot(sst_c_mask[[1]], main="SST mask")
plot(sal_psu_mask[[1]], main="Salinity mask")
plot(o2_molm3_mask[[1]], main="Oxygen mask")
plot(chl_mgm3_meansubsd_mask[[1]], main="Chlorophyll mask")
# Read EEZs
eezdir <- "data/eezs"
eezs <- raster(file.path(eezdir, "eezs_v10_raster_10km.tif"))
eez_key <- read.csv(file.path(eezdir, "eezs_v10_key.csv"), as.is=T)
# Build EEZ mask
eez_mask <- eezs
eez_mask[!is.na(eez_mask)] <- 1
# Read climate forecasts
climatedir <- "data/climate/data/gfdl/GFDL-ESM2G/4rasters_scaled"
ras_sst_c_min <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_tos_degC_annual_min_scaled.tif")))
ras_sst_c_max <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_tos_degC_annual_max_scaled.tif")))
ras_sal_psu_mean <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_so_psu_annual_mean_scaled.tif")))
ras_o2_molm3_mean <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_o2_mol_m3_annual_mean_scaled.tif")))
ras_chl_mgm3_meansubsd <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_chl_mg_m3_annual_mean_minus_sd_scaled.tif")))
ras_arag_sat_mean <- raster::brick(file.path(climatedir, paste0("GFDL_ESM2M_", rcp, "_arag_sat_annual_mean_scaled.tif")))
ras_layer_names <- readRDS(file.path(climatedir, "GFDL_ESM2M_rcp85_arag_sat_annual_mean_scaled_layer_names.rds"))
# Add raster layer names
names(ras_sst_c_min) <- names(ras_sst_c_max) <- names(ras_sal_psu_mean) <- names(ras_o2_molm3_mean) <-
names(ras_chl_mgm3_meansubsd) <- names(ras_arag_sat_mean) <- ras_layer_names
# Check rasters
env_ras_check <- compareRaster(eezs, ras_sst_c_min, ras_sst_c_max, ras_sal_psu_mean, ras_o2_molm3_mean, ras_chl_mgm3_meansubsd, ras_arag_sat_mean)
if(env_ras_check==F){print("EEZ and climate forecast rasters DO NOT have the same projection, extent, and resolution.")}
# 1. Extract parameters
####################################
# Growth and harvest parameters
spp <- species$species
type <- species$type
linf_cm <- species$linf_cm
k <- species$k
harvest_cm <- species$harvest_cm
harvest_g <- species$harvest_g
harvest_yr <- species$harvest_yr
a <- species$a
b <- species$b
fcr <- species$fcr
price_usd_mt <- species$price_usd_mt_isscaap
print(spp)
# Environmental tolerance parameters
sst_c_min <- species$sst_c_min
sst_c_max <- species$sst_c_max
sal_psu_min <- species$sal_psu_min
sal_psu_max <- species$sal_psu_max
if(type=="Finfish"){
o2_molm3_min <- 0.1378
chl_mgm3_meansubsd_min <- NA
arag_sat_min <- NA
}
if(type=="Bivalve"){
o2_molm3_min <- 0.0622
chl_mgm3_meansubsd_min <- 0.4
arag_sat_min <- 1
}
# 2. Calculate harvest/sqkm/yr
####################################
# Finfish farm design
if(type=="Finfish"){
# Finfish farm design
farm_design <- tibble(type="finfish",
area_sqkm=1,
ncages=24,
cage_vol_m3=9000,
juv_m3=20) %>%
mutate(nstocked=ncages * cage_vol_m3 * juv_m3)
}
# Bivalve farm design
if(type=="Bivalve"){
farm_design <- tibble(type="bivalve",
area_sqkm=1,
nlines=100,
line_m=4000,
juv_ft=100) %>%
mutate(line_ft=measurements::conv_unit(line_m, "m", "ft"),
nstocked=nlines * line_ft * juv_ft)
}
# Calculate yield per year (kg/yr) for 1 sqkm farm
farm_kg_yr <- harvest_g/1000 * farm_design$nstocked / harvest_yr
farm_mt_yr <- farm_kg_yr / 1000
# 3. Identify viable cells
####################################
# Years to evaluate
yrs_available <- as.numeric(ras_layer_names)
yrs_do_indices <- which(yrs_available %in% years)
# Viable cells
print("... identifying viable cells")
sst_c_mask <- ras_sst_c_min[[yrs_do_indices]] >= sst_c_min & ras_sst_c_max[[yrs_do_indices]] <= sst_c_max
sal_psu_mask <- ras_sal_psu_mean[[yrs_do_indices]] >= sal_psu_min & ras_sal_psu_mean[[yrs_do_indices]] <= sal_psu_max
# Bivalve
if(type=="Bivalve"){
o2_molm3_mask <- ras_o2_molm3_mean[[yrs_do_indices]] >= o2_molm3_min
chl_mgm3_meansubsd_mask <- ras_chl_mgm3_meansubsd[[yrs_do_indices]] >= chl_mgm3_meansubsd_min
arag_sat_mask <- ras_arag_sat_mean[[yrs_do_indices]] >= arag_sat_min
vcells <- eez_mask * sst_c_mask * sal_psu_mask * o2_molm3_mask * chl_mgm3_meansubsd_mask * arag_sat_mask
NAvalue(vcells) <- 0
}
# Finfish
if(type=="Finfish"){
o2_molm3_mask <- ras_o2_molm3_mean[[yrs_do_indices]] >= o2_molm3_min
vcells <- eez_mask * sst_c_mask * sal_psu_mask * o2_molm3_mask
NAvalue(vcells) <- 0
}
plot(vcells[[1]], main="Suitable cells")
spp
# Clear workspace
rm(list = ls())
# Setup
################################################################################
# Packages
library(raster)
library(ggplot2)
library(tidyverse)
# Directories
codedir <- "code"
sppdir <- "data/species/data"
outputdir <- "output/raw"
# Read aquacast function
source(file.path(codedir, "aquacast.R"))
source(file.path(codedir, "calc_costs.R"))
# Read species data
load(file.path(sppdir, "aquaculture_species_key.Rdata"))
# Setup for testing
################################################################################
# Check error in following species:
# Crassostrea rhizophorae, Crassostrea tulipa, Crassostrea virginica, Ostrea chilensis:  Error in value[j, ] : incorrect number of dimensions
# Format data
data <- data %>%
mutate(type=recode(class,
"Bivalvia"="Bivalve",
"Actinopterygii"="Finfish"))
# Loop through species and run model
i <- 1
for(i in 1:nrow(data)){
# for(i in 1:nrow(bivalves)){
# Parameters
species <- data[i,]
# years <- 2021
years <- c(2021, 2051, 2100)
# Forecast aquaculture potential
output <- aquacast(species=species, years=years, rcp="rcp85", outdir=outputdir, plot=T)
}
# Clear
rm(list = ls())
# Setup
#####################################################################################
# Packages
library(sf)
library(rgeos)
library(raster)
library(tidyverse)
library(ggplot2)
library(rnaturalearth)
library(rnaturalearthdata)
library(fasterize)
# Directories
tempdir <- "data/template"
costdir <-  "data/costs/data"
eezdir <- "data/eezs"
# Read raster template
ras_temp <- raster(file.path(tempdir, "world_raster_template_10km.tif"))
values(ras_temp) <- 1:ncell(ras_temp)
# Read EEZ raster and key
eezs <- raster(file.path(eezdir, "eezs_v10_raster_10km.tif"))
eezs_sf <- readRDS(file.path(eezdir, "eezs_v10_polygons.Rds"))
eez_key <- read.csv(file.path(eezdir, "eezs_v10_key.csv"), as.is=T)
compareRaster(eezs, ras_temp)
# Get coastline
# coast1 <- rnaturalearth::ne_coastline(scale="small", returnclass="sf") %>% sf::st_transform(crs(ras_temp)) %>% sf::as_Spatial() # most coarse
# coast2 <- rnaturalearth::ne_coastline(scale="medium", returnclass="sf") %>% sf::st_transform(crs(ras_temp))
# coast3 <- rnaturalearth::ne_coastline(scale="large", returnclass="sf") %>% sf::st_transform(crs(ras_temp)) # most detailed
# Get most countries
most <- rnaturalearth::ne_countries(scale="large", returnclass="sf") %>% sf::st_transform(crs(ras_temp))
most <- rnaturalearth::ne_countries(scale="large", returnclass="sf") %>% sf::st_transform(crs(ras_temp))
# Get and buffer tiny countries
tiny <- rnaturalearth::ne_countries(type="tiny_countries", returnclass="sf") %>% sf::st_transform(crs(ras_temp))
tiny_2km <- sf::st_buffer(tiny, dist=2000) %>%
sf::st_cast("MULTIPOLYGON")
# Merge
land <- rbind(most %>% select(iso_a3, geometry), tiny_2km %>% select(iso_a3, geometry))
?rasterize
land_ras <- rasterize(x=land, y=ras_temp, background=NA, getCover=T)
# Distance from land
cdist_m <- raster::distance(land_ras)
cdist_km <- cdist_m / 1000
# Plot distance from land
plot(cdist_km)
# Check against template raster
compareRaster(cdist_km, ras_temp)
land_ras
hist(land_ras)
table(getValues(land_ras))
reclassify
?reclassify
land_ras_rec <- reclassify(land_ras, rcl=matrix(data=c(0,1,1), ncol=3))
land_ras_rec
table(getValues(land_ras_rec))
NAvalue(land_ras_rec)
land_ras_rec[land_ras_rec==0] <- NA
# Distance from land
cdist_m <- raster::distance(land_ras)
cdist_km <- cdist_m / 1000
# Plot distance from land
plot(cdist_km)
land_ras_rec[land_ras_rec==0]
land_ras_rec <- reclassify(land_ras, rcl=matrix(data=c(0,1,1), ncol=3))
land_ras_rec[land_ras_rec==0]
land_ras_rec[land_ras_rec==0] <- NA
# Distance from land
cdist_m <- raster::distance(land_ras_rec)
cdist_km <- cdist_m / 1000
# Plot distance from land
plot(cdist_km)
# Check against template raster
compareRaster(cdist_km, ras_temp)
# Export
writeRaster(cdist_km, filename=file.path(costdir, "dist_to_coast_km_10km_inhouse.grd"), format="raster", overwrite=T)
# Setup
#####################################################################################
# Packages
library(raster)
library(tidyverse)
library(countrycode)
library(rnaturalearth)
# Directories
eezdir <- "data/eezs"
tempdir <- "data/template"
costdir <- "data/costs/data"
plotdir <- "data/costs/figures"
# Read raster template
ras_temp <- raster(file.path(tempdir, "world_raster_template_10km.tif"))
# Read EEZ raster and key
eezs <- raster(file.path(eezdir, "eezs_v10_raster_10km.tif"))
eezs_sf <- readRDS(file.path(eezdir, "eezs_v10_polygons.Rds"))
eez_key <- read.csv(file.path(eezdir, "eezs_v10_key.csv"), as.is=T)
compareRaster(eezs, ras_temp)
# World layer
world <- rnaturalearth::ne_countries(scale = "large", type = "countries", returnclass = "sf")
# Distance to shore (m) raster
cdist_km <- raster(file.path(costdir, "dist_to_coast_km_10km_inhouse.grd"))
# WB data
wb_costs <- read.csv(file.path(costdir, "eez_wb_diesel_income_costs.csv"), as.is=T)
# Number of farms per cell
#####################################################################################
# Number of farms per cells
cell_sqkm <- prod(res(eezs)/1000)
nfarms <- cell_sqkm
# 1. Calculate fuel costs
#####################################################################################
# Fuel cost parameters
# From Lester et al. 2018
vessel_n <- 2
vessel_lph <- 60.6 # vessel fuel efficiency, liters per hour (from 16 g/hr in Lester et al. 2018)
vessel_kmh <- 12.9 # vessel speed, km per hour (from 8 mph in Lester et al. 2018)
vessel_trips_yr <- 416 # two identical vessels - one making 5 trips/week and the other 3 trips/week (from Lester et al. 2018)
# Create fuel price raster
fuel_usd_l_ras <- reclassify(eezs, rcl=select(wb_costs, eez_code, diesel_usd_l))
# plot(fuel_usd_l_ras)
# Fuel cost raster
fuel_usd_yr_farm <- (2 * cdist_km) / vessel_kmh * vessel_lph * fuel_usd_l_ras * vessel_trips_yr
# Quick plot
plot(fuel_usd_yr_farm/1e6, main="Annual fuel cost per farm (USD millions)", xaxt="n", yaxt="n",
col=freeR::colorpal(rev(RColorBrewer::brewer.pal(n=9, name="RdYlBu")), 50))
# Export data
writeRaster(fuel_usd_yr_farm, file.path(costdir, "fuel_cost_per_farm.tif"), overwrite=T)
# Labor cost parameters
# From Lester et al. 2018
worker_n <- 8 # 8 workers per farm
worker_hrs <- 2080 # 40 hour weeks x 52 weeks / year = 2080 hours / year (does not include transport time)
# Create wages raster
wb_costs$wages_usd_hr <- wb_costs$income_usd_yr / worker_hrs
wages_usd_hr_ras <- reclassify(eezs, rcl=select(wb_costs, eez_code, wages_usd_hr))
# Quick plot
plot(wages_usd_hr_ras, main="Worker wages (USD per hour)", xaxt="n", yaxt="n",
col=freeR::colorpal(RColorBrewer::brewer.pal(n=9, name="RdYlBu"), 50))
# Calculate number of transit hours
transit_hrs_ras <- (2 * cdist_km) / vessel_kmh * vessel_trips_yr
# Calculate annnual wage cost per farm
wages_usd_yr_farm <- wages_usd_hr_ras * worker_n * (worker_hrs + transit_hrs_ras)
# Quick plot
plot(wages_usd_yr_farm/1e6, main="Annual wage cost per farm (USD millions)", xaxt="n", yaxt="n",
col=freeR::colorpal(rev(RColorBrewer::brewer.pal(n=9, name="RdYlBu")), 50))