2. Combine dsclim data from past to future
Source:vignettes/a2_dsclim_past_to_future.Rmd
a2_dsclim_past_to_future.RmdLoading past data into R environment
It is crucial to ensure a seamless transition between data from different sources when utilising climate data for past and future periods. This is important even when the overlapping period data source is the same. We have illustrated this concept through a graphical representation in this vignette. To exemplify, we will load the downscaled past climate data into R and extract a 30-year period (i.e., 1960-1990) in two locations.
library(dsclimtools)
library(stars)
library(units)
library(ggplot2)
point1 <- c(0, 41.5) %>% matrix(1, 2) %>% sf::st_point(dim = "XYZ")
point2 <- c(0, 42.8) %>% matrix(1, 2) %>% sf::st_point(dim = "XYZ")
past1 <- dsclimtools::read_dsclim("../data/dsclim",
var = "tas",
y_start = 10,
y_end = 40,
calendar_dates = TRUE,
sf = point1,
proxy = FALSE)
past2 <- dsclimtools::read_dsclim("../data/dsclim",
var = "tas",
y_start = 10,
y_end = 40,
calendar_dates = TRUE,
sf = point2,
proxy = FALSE)
figure <- ggplot() +
geom_line(data = drop_units(as.data.frame(past1)),
aes(x = time,
y = tas),
color = "#E69F00",
alpha = 0.3) +
geom_line(data = drop_units(as.data.frame(aggregate(past1,
by = calendar_dates(10, 41, "1 year"),
FUN = mean,
na.rm = TRUE))),
aes(x = time,
y = tas),
color = "#E69F00") +
geom_line(data = drop_units(as.data.frame(past2)),
aes(x = time,
y = tas),
color = "#56B4E9",
alpha = 0.3) +
geom_line(data = drop_units(as.data.frame(aggregate(past2,
by = calendar_dates(10, 41, "1 year"),
FUN = mean,
na.rm = TRUE))),
aes(x = time,
y = tas),
color = "#56B4E9")
Figure 1. Temperatures for two points of the
study area during the calibration period (1960-1990)
Extracting future data in point locations for different climate change scenarios.
In this section, we present future climate data extracted from the same point locations for the various climate change scenarios. Furthermore, the data is aggregated on an annual basis in order to simplify any potential oscillations and provide a clearer visual representation.
rcps <- c("rcp2.6", "rcp4.5", "rcp6.0", "rcp8.5")
gcms <- c("CESM1-CAM5", "CSIRO-Mk3-6-0", "IPSL-CM5A-MR")
df <- expand.grid(gcms, rcps)
get_dscmip_point_data <- function(folder, var, start, end, rcp, gcm, sf, agg = NULL){
data <- dsclimtools::read_dsclim(folder,
var,
start,
end,
rcp,
gcm,
calendar_dates = TRUE,
sf,
proxy = FALSE)
if(!is.null(agg)){
data <- aggregate(data,
by = calendar_dates(start,
end,
agg),
FUN = mean,
na.rm = TRUE)
}
data <- as.data.frame(data)
data$gcm <- gcm
data$rcp <- rcp
data
}
fut1 <- mapply(get_dscmip_point_data,
rcp = df$Var2,
gcm = df$Var1,
MoreArgs = list(folder = "../data/dsclim",
var = "tas",
start = 41,
end = 150,
sf = point1),
SIMPLIFY = FALSE)
fut1 <- do.call(rbind, fut1)
fut1.agg <- mapply(get_dscmip_point_data,
rcp = df$Var2,
gcm = df$Var1,
MoreArgs = list(folder = "../data/dsclim",
var = "tas",
start = 41,
end = 150,
sf = point1,
agg = "1 year"),
SIMPLIFY = FALSE)
fut1.agg <- do.call(rbind, fut1.agg)
fut2 <- mapply(get_dscmip_point_data,
rcp = df$Var2,
gcm = df$Var1,
MoreArgs = list(folder = "../data/dsclim",
var = "tas",
start = 41,
end = 150,
sf = point2),
SIMPLIFY = FALSE)
fut2 <- do.call(rbind, fut2)
fut2.agg <- mapply(get_dscmip_point_data,
rcp = df$Var2,
gcm = df$Var1,
MoreArgs = list(folder = "../data/dsclim",
var = "tas",
start = 41,
end = 150,
sf = point2,
agg = "1 year"),
SIMPLIFY = FALSE)
fut2.agg <- do.call(rbind, fut2.agg)
figure <- figure +
stat_summary(data = drop_units(fut1),
aes(x = time,
y = tas,
group = rcp,
color = rcp),
fun = mean,
geom = 'line',
alpha = 0.3) +
stat_summary(data = drop_units(fut1.agg),
aes(x = time,
y = tas,
group = rcp,
color = rcp),
fun = mean,
geom = 'line') +
stat_summary(data = drop_units(fut2),
aes(x = time,
y = tas,
group = rcp,
color = rcp),
fun = mean,
geom = 'line',
alpha = 0.3) +
stat_summary(data = drop_units(fut2.agg),
aes(x = time,
y = tas,
group = rcp,
color = rcp),
fun = mean,
geom = 'line') +
theme_light() +
labs(y = "Surface Temperature (ºC)",
x = "Time (calendar years)") +
coord_cartesian(xlim = as.Date(c('1980-01-01','2100-12-31'))) +
scale_color_discrete(name = "Emission\npathways",
guide = "legend")
figure
Figure 2. Temperatures for the same two points
of the study area for the calibration period (1960-1990) and the average
future projections among three climate models (e.g. CESM1-CAM5,
CSIRO-Mk3-6-0, and IPSL-CM5A-MR) under four different Representative
Concentration Pathways (RCPs; rcp2.6, rcp4.5, rcp6.0, rcp8.5).
The objective of this plot is to demonstrate the gradual transition between past and future data, as well as the divergence between the various climate change scenarios.