Last active
November 13, 2018 15:55
-
-
Save TonyLadson/066a97f5be10d3635067a57e12b3d0cc to your computer and use it in GitHub Desktop.
Recipe for creating a flood frequency plot see https://tonyladson.wordpress.com/2017/06/09/flood-frequency-plots-using-ggplot/
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
library(scales) | |
library(tidyverse) | |
library(stringr) | |
# The tribble function provides a good way of entering small data sets | |
# flows are in ML/d | |
ff_comparison <- tribble( | |
~aep, ~broken_casey, ~seven_kialla, | |
0.5, 6900, NA, | |
0.2, 23450, 21400, | |
0.1, 37800, 33400, | |
0.05, 53600, 46300, | |
0.02, 76000, 64100, | |
0.01, 95000, 77700 | |
) | |
# Add a column of normal quantiles that correspond to | |
# AEP values and a column of 1 in X AEP values | |
ff_comparison <- ff_comparison %>% | |
mutate(z = qnorm(1 - aep), | |
aep_1inX = 1/aep) | |
# Copy this dataframe so we can use it for labelling | |
my_labels <- ff_comparison | |
my_labels | |
# Change the format of the dataframe to make plotting easier | |
ff_comparison <- ff_comparison %>% | |
gather(key = 'source', value = peak_flow, -aep, -z, -aep_1inX) # one observation per row | |
# Plot | |
ff_comparison %>% | |
mutate(peak_flow = peak_flow/86.4) %>% # convert to cumec | |
ggplot(aes(x = z, y = peak_flow, colour = source)) + | |
geom_point() + | |
geom_line() + | |
scale_y_continuous(name = expression(Peak~flow~(m^{3}~s^{-1})), labels = scales::comma, | |
breaks = seq(0, 1000, 200)) + | |
# scale_y_log10(name = expression(Peak~flow~(m^{3}~s^{-1})), labels = comma, # log scale | |
# breaks = c(seq(10000, 100000, 10000))) + | |
scale_x_continuous(name = 'AEP', breaks = my_labels$z, labels = str_c(100* my_labels$aep,'%'), | |
sec.axis = dup_axis(name = 'AEP (1 in X years)', labels = my_labels$aep_1inX)) + | |
scale_colour_manual(name = 'Waterway', labels = c('Broken River at Casey Weir', 'Sevens Cks at Kialla West'), | |
values = c('light blue', 'dark blue')) + | |
theme_grey(base_size = 12) + | |
theme(legend.position = c(0.8, 0.2)) | |
# Can also use the Probit transformation to achieve much the same result | |
# Note trans = 'probit' has been added to scale_x_continuous and the x aesthetic is (1 - aep) | |
# We need the 1 - aep for the exceedance probability | |
ff_comparison %>% | |
mutate(peak_flow = peak_flow/86.4) %>% # convert to cumec | |
ggplot(aes(x = 1-aep, y = peak_flow, colour = source)) + | |
geom_point() + | |
geom_line() + | |
scale_y_continuous(name = expression(Peak~flow~(m^{3}~s^{-1})), labels = scales::comma, | |
breaks = seq(0, 1000, 200)) + | |
# scale_y_log10(name = expression(Peak~flow~(m^{3}~s^{-1})), labels = comma, # log scale | |
# breaks = c(seq(10000, 100000, 10000))) + | |
scale_x_continuous(name = 'AEP', | |
trans = 'probit', | |
breaks = 1-my_labels$aep, | |
labels = str_c(100* my_labels$aep,'%'), | |
sec.axis = dup_axis(name = 'AEP (1 in X years)', | |
labels = my_labels$aep_1inX)) + | |
scale_colour_manual(name = 'Waterway', labels = c('Broken River at Casey Weir', 'Sevens Cks at Kialla West'), | |
values = c('light blue', 'dark blue')) + | |
theme_grey(base_size = 12) + | |
theme(legend.position = c(0.8, 0.2)) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment