In tito-am/prsim.tools: Librairie complémentaire à PRSIM. Cette librarie permet de simuler, vérifier et créer les figures pour l'analyse des séries stochastiques.
Introduction
This is a notebook intended to explain at large what we have done with PRSIM on the Outaouais River Watershed. First of all, let's get an idea of the area of study.
The watershed surface is about 140 000 $km^{2}$ and consists of 42 basins. On the Outaouais river watershed we can find infrastructures that are designed to withstand PMF, 1:10 000 and 1:1 000 return year floods depending on the canadian dam classification.
Our main interest in PRSIM is the fact that spatial and temporal intercorrelations are taken into account while producing the stochastic simulations.
Extreme floods at the basin level
In our first trials with PRSIM we found out that for some basins certain unrealistic peakflows where simulated. Even for extreme cases. In fact, some 1:100 year peakflows where higher than our PMF (Probable maximum flood). In order to continue to explore the generation, we limited the kappa generation function to 0.9995 right-side limit. With this modification, we found out more realistic hydrogram shapes and the statistics up to 1:100 years return peakflow was coherent with what we calculated using the observations. is this a correct procedure?
As said before, in order to explore our simulations we first calculated the marginal return periods at each basin of the Outaouais following the cunnane plotting positions and compare them to our inferences on the observed data.
The next table is our simulated peakflow during spring for return periods up to 10 000 year flood:
library(xlsx)
library(kableExtra)
file<-'/home/tito/Documents/github/prsim.tools/data/afl_series_obs_prsim_volume_pointe_1960-2017_vers2(1).xlsx'
dt<-read.xlsx(file, 'rmarkdown_prt_volume', header=TRUE, colClasses=NA)
dt %>%
kbl() %>%
kable_paper("hover", full_width = F)
As you can see, in some watersheds the right tail is pretty 'flat'. See at Rapide-2. Meaning that the difference in between the extreme quantiles is not very large.
The next table is our simulated 70-day volume during spring for return periods up to 10 000 year flood:
library(xlsx)
library(kableExtra)
file<-'/home/tito/Documents/github/prsim.tools/data/afl_series_obs_prsim_volume_pointe_1960-2017_vers2(1).xlsx'
dt<-read.xlsx(file, 'rmarkdown_prt_peakflow', header=TRUE, colClasses=NA)
dt %>%
kbl() %>%
kable_paper("hover", full_width = F)
The next table is our simulated peakflow volume during summer-autumn for return periods up to 10 000 year flood:
library(xlsx)
library(kableExtra)
file<-'/home/tito/Documents/github/prsim.tools/data/afl_series_obs_prsim_volume_pointe_1960-2017_vers2(1).xlsx'
dt<-read.xlsx(file, 'rmarkdown_ete_peakflow', header=TRUE, colClasses=NA)
dt %>%
kbl() %>%
kable_paper("hover", full_width = F)
Summary statistics at the basin level
As you will see in our summary statistic analyses, sometimes the observations entries show higher peaks than the simulations. Specially during summer. This time again, it's Rapide-2.
Next, I compared the extent of the 170 000 prsim simulations light green font and also the mean green line. The observations forced into PRSIM (57 years) are in red. I plotted the maximum, the mean and the minimum.
I will show some of the 43 basins. As you can see, most of the means seem to be well simulated.
Gatineau
library(plotly)
library(readr)
library(tidyverse)
#
load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata')
Bask<-as.data.frame(tests['Baskatong'])
Qobs_prsim_mean<-Bask %>% group_by(Baskatong.MM,Baskatong.DD) %>%summarise(AvgQ=mean(Baskatong.Qobs,na.rm = TRUE))
Qobs_prsim_max<-Bask %>% group_by(Baskatong.MM,Baskatong.DD) %>%summarise(MaxQ=max(Baskatong.Qobs,na.rm = TRUE))
Qobs_prsim_min<-Bask %>% group_by(Baskatong.MM,Baskatong.DD) %>%summarise(MinQ=min(Baskatong.Qobs,na.rm = TRUE))
#Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019)
path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/'
fichiers<-list.files(path,pattern = '.csv')
Q10000_etude2019<-read_csv(paste0(path,fichiers[1]))
jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day")
#Statistiques sommaires PRSIM
load('/home/tito/Baskatong.Rdata')
df_final_clean<-df_final[!duplicated(df_final$julian_day),]
x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day")
random_y <- df_final_clean$MaxQ
data <- data.frame(x, random_y)
trace1<-df_final_clean$AvgQ
trace2<-df_final_clean$MinQ
trace3<-Qobs_prsim_mean$AvgQ
trace4<-Qobs_prsim_max$MaxQ
trace5<-Qobs_prsim_min$MinQ
fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% layout(title = "Average, High and Low PRSIM at Baskatong",
paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)',
xaxis = list(title = "Days",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE),
yaxis = list(title = "Streamflow (cms)",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE))
fig
Outaouais Superieur
library(plotly)
library(readr)
library(tidyverse)
#
load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata')
Bask<-as.data.frame(tests['Dozois'])
Qobs_prsim_mean<-Bask %>% group_by(Dozois.MM,Dozois.DD) %>%summarise(AvgQ=mean(Dozois.Qobs,na.rm = TRUE))
Qobs_prsim_max<-Bask %>% group_by(Dozois.MM,Dozois.DD) %>%summarise(MaxQ=max(Dozois.Qobs,na.rm = TRUE))
Qobs_prsim_min<-Bask %>% group_by(Dozois.MM,Dozois.DD) %>%summarise(MinQ=min(Dozois.Qobs,na.rm = TRUE))
#Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019)
path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/'
fichiers<-list.files(path,pattern = '.csv')
Q10000_etude2019<-read_csv(paste0(path,fichiers[1]))
jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day")
#Statistiques sommaires PRSIM
load('/home/tito/Dozois.Rdata')
df_final_clean<-df_final[!duplicated(df_final$julian_day),]
x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day")
random_y <- df_final_clean$MaxQ
data <- data.frame(x, random_y)
trace1<-df_final_clean$AvgQ
trace2<-df_final_clean$MinQ
trace3<-Qobs_prsim_mean$AvgQ
trace4<-Qobs_prsim_max$MaxQ
trace5<-Qobs_prsim_min$MinQ
fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% layout(title = "Average, High and Low PRSIM at Dozois",
paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)',
xaxis = list(title = "Days",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE),
yaxis = list(title = "Streamflow (cms)",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE))
fig
library(plotly)
library(readr)
library(tidyverse)
#
load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata')
Bask<-as.data.frame(tests['Lac Victoria et lac Granet'])
Qobs_prsim_mean<-Bask %>% group_by(Lac.Victoria.et.lac.Granet.MM,Lac.Victoria.et.lac.Granet.DD) %>%summarise(AvgQ=mean(Lac.Victoria.et.lac.Granet.Qobs,na.rm = TRUE))
Qobs_prsim_max<-Bask %>% group_by(Lac.Victoria.et.lac.Granet.MM,Lac.Victoria.et.lac.Granet.DD) %>%summarise(MaxQ=max(Lac.Victoria.et.lac.Granet.Qobs,na.rm = TRUE))
Qobs_prsim_min<-Bask %>% group_by(Lac.Victoria.et.lac.Granet.MM,Lac.Victoria.et.lac.Granet.DD) %>%summarise(MinQ=min(Lac.Victoria.et.lac.Granet.Qobs,na.rm = TRUE))
#Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019)
path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/'
fichiers<-list.files(path,pattern = '.csv')
Q10000_etude2019<-read_csv(paste0(path,fichiers[1]))
jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day")
#Statistiques sommaires PRSIM
load('/home/tito/Lac Victoria et lac Granet.Rdata')
df_final_clean<-df_final[!duplicated(df_final$julian_day),]
x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day")
random_y <- df_final_clean$MaxQ
data <- data.frame(x, random_y)
trace1<-df_final_clean$AvgQ
trace2<-df_final_clean$MinQ
trace3<-Qobs_prsim_mean$AvgQ
trace4<-Qobs_prsim_max$MaxQ
trace5<-Qobs_prsim_min$MinQ
fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% layout(title = "Average, High and Low PRSIM at Lac Victoria et lac Granet",
paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)',
xaxis = list(title = "Days",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE),
yaxis = list(title = "Streamflow (cms)",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE))
fig
library(plotly)
library(readr)
library(tidyverse)
#
load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata')
Bask<-as.data.frame(tests['Rapide-2'])
Qobs_prsim_mean<-Bask %>% group_by(Rapide.2.MM,Rapide.2.DD) %>%summarise(AvgQ=mean(Rapide.2.Qobs,na.rm = TRUE))
Qobs_prsim_max<-Bask %>% group_by(Rapide.2.MM,Rapide.2.DD) %>%summarise(MaxQ=max(Rapide.2.Qobs,na.rm = TRUE))
Qobs_prsim_min<-Bask %>% group_by(Rapide.2.MM,Rapide.2.DD) %>%summarise(MinQ=min(Rapide.2.Qobs,na.rm = TRUE))
#Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019)
path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/'
fichiers<-list.files(path,pattern = '.csv')
Q10000_etude2019<-read_csv(paste0(path,fichiers[1]))
jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day")
#Statistiques sommaires PRSIM
load('/home/tito/Rapide-2.Rdata')
df_final_clean<-df_final[!duplicated(df_final$julian_day),]
x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day")
random_y <- df_final_clean$MaxQ
data <- data.frame(x, random_y)
trace1<-df_final_clean$AvgQ
trace2<-df_final_clean$MinQ
trace3<-Qobs_prsim_mean$AvgQ
trace4<-Qobs_prsim_max$MaxQ
trace5<-Qobs_prsim_min$MinQ
fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% layout(title = "Average, High and Low PRSIM at Rapide-2",
paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)',
xaxis = list(title = "Days",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE),
yaxis = list(title = "Streamflow (cms)",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE))
fig
library(plotly)
library(readr)
library(tidyverse)
#
load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata')
Bask<-as.data.frame(tests['Rapide-7'])
Qobs_prsim_mean<-Bask %>% group_by(Rapide.7.MM,Rapide.7.DD) %>%summarise(AvgQ=mean(Rapide.7.Qobs,na.rm = TRUE))
Qobs_prsim_max<-Bask %>% group_by(Rapide.7.MM,Rapide.7.DD) %>%summarise(MaxQ=max(Rapide.7.Qobs,na.rm = TRUE))
Qobs_prsim_min<-Bask %>% group_by(Rapide.7.MM,Rapide.7.DD) %>%summarise(MinQ=min(Rapide.7.Qobs,na.rm = TRUE))
#Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019)
path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/'
fichiers<-list.files(path,pattern = '.csv')
Q10000_etude2019<-read_csv(paste0(path,fichiers[1]))
jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day")
#Statistiques sommaires PRSIM
load('/home/tito/Rapide-7.Rdata')
df_final_clean<-df_final[!duplicated(df_final$julian_day),]
x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day")
random_y <- df_final_clean$MaxQ
data <- data.frame(x, random_y)
trace1<-df_final_clean$AvgQ
trace2<-df_final_clean$MinQ
trace3<-Qobs_prsim_mean$AvgQ
trace4<-Qobs_prsim_max$MaxQ
trace5<-Qobs_prsim_min$MinQ
fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% layout(title = "Average, High and Low PRSIM at Rapide-7",
paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)',
xaxis = list(title = "Days",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE),
yaxis = list(title = "Streamflow (cms)",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE))
fig
library(plotly)
library(readr)
library(tidyverse)
#
load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata')
Bask<-as.data.frame(tests['Lac des Quinze'])
Qobs_prsim_mean<-Bask %>% group_by(Lac.des.Quinze.MM,Lac.des.Quinze.DD) %>%summarise(AvgQ=mean(Lac.des.Quinze.Qobs,na.rm = TRUE))
Qobs_prsim_max<-Bask %>% group_by(Lac.des.Quinze.MM,Lac.des.Quinze.DD) %>%summarise(MaxQ=max(Lac.des.Quinze.Qobs,na.rm = TRUE))
Qobs_prsim_min<-Bask %>% group_by(Lac.des.Quinze.MM,Lac.des.Quinze.DD) %>%summarise(MinQ=min(Lac.des.Quinze.Qobs,na.rm = TRUE))
#Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019)
path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/'
fichiers<-list.files(path,pattern = '.csv')
Q10000_etude2019<-read_csv(paste0(path,fichiers[1]))
jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day")
#Statistiques sommaires PRSIM
load('/home/tito/Lac des Quinze.Rdata')
df_final_clean<-df_final[!duplicated(df_final$julian_day),]
x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day")
random_y <- df_final_clean$MaxQ
data <- data.frame(x, random_y)
trace1<-df_final_clean$AvgQ
trace2<-df_final_clean$MinQ
trace3<-Qobs_prsim_mean$AvgQ
trace4<-Qobs_prsim_max$MaxQ
trace5<-Qobs_prsim_min$MinQ
fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE)
fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter')
fig <- fig %>% layout(title = "Average, High and Low PRSIM at Rapide-7",
paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)',
xaxis = list(title = "Days",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE),
yaxis = list(title = "Streamflow (cms)",
gridcolor = 'rgb(255,255,255)',
showgrid = TRUE,
showline = FALSE,
showticklabels = TRUE,
tickcolor = 'rgb(127,127,127)',
ticks = 'outside',
zeroline = FALSE))
fig
Future work
- Increase the probability of generation (Kappa)
- Do seasonality analysis (Boxplots per season as in Manuela's vignettes)
- Cross-correlation
- PACF, ACf (Replicate Manuela's vignettes with our 170 000 simulations)
- Madogram (Explore Pierre Ribatet package. Try to replicate P-Madogram in Manuela's article)
Questions
- What to do with Rapide-2 type of case?
- Could you please explain more in depth the interpretation of PACF, ACF and P-Madogram?
Thank you!
tito-am/prsim.tools documentation built on March 17, 2021, 9:09 p.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.
Introduction
This is a notebook intended to explain at large what we have done with PRSIM on the Outaouais River Watershed. First of all, let's get an idea of the area of study.
The watershed surface is about 140 000 $km^{2}$ and consists of 42 basins. On the Outaouais river watershed we can find infrastructures that are designed to withstand PMF, 1:10 000 and 1:1 000 return year floods depending on the canadian dam classification.
Our main interest in PRSIM is the fact that spatial and temporal intercorrelations are taken into account while producing the stochastic simulations.
Extreme floods at the basin level
In our first trials with PRSIM we found out that for some basins certain unrealistic peakflows where simulated. Even for extreme cases. In fact, some 1:100 year peakflows where higher than our PMF (Probable maximum flood). In order to continue to explore the generation, we limited the kappa generation function to 0.9995 right-side limit. With this modification, we found out more realistic hydrogram shapes and the statistics up to 1:100 years return peakflow was coherent with what we calculated using the observations. is this a correct procedure?
As said before, in order to explore our simulations we first calculated the marginal return periods at each basin of the Outaouais following the cunnane plotting positions and compare them to our inferences on the observed data.
The next table is our simulated peakflow during spring for return periods up to 10 000 year flood:
library(xlsx) library(kableExtra) file<-'/home/tito/Documents/github/prsim.tools/data/afl_series_obs_prsim_volume_pointe_1960-2017_vers2(1).xlsx' dt<-read.xlsx(file, 'rmarkdown_prt_volume', header=TRUE, colClasses=NA) dt %>% kbl() %>% kable_paper("hover", full_width = F)
As you can see, in some watersheds the right tail is pretty 'flat'. See at Rapide-2. Meaning that the difference in between the extreme quantiles is not very large.
The next table is our simulated 70-day volume during spring for return periods up to 10 000 year flood:
library(xlsx) library(kableExtra) file<-'/home/tito/Documents/github/prsim.tools/data/afl_series_obs_prsim_volume_pointe_1960-2017_vers2(1).xlsx' dt<-read.xlsx(file, 'rmarkdown_prt_peakflow', header=TRUE, colClasses=NA) dt %>% kbl() %>% kable_paper("hover", full_width = F)
The next table is our simulated peakflow volume during summer-autumn for return periods up to 10 000 year flood:
library(xlsx) library(kableExtra) file<-'/home/tito/Documents/github/prsim.tools/data/afl_series_obs_prsim_volume_pointe_1960-2017_vers2(1).xlsx' dt<-read.xlsx(file, 'rmarkdown_ete_peakflow', header=TRUE, colClasses=NA) dt %>% kbl() %>% kable_paper("hover", full_width = F)
Summary statistics at the basin level
As you will see in our summary statistic analyses, sometimes the observations entries show higher peaks than the simulations. Specially during summer. This time again, it's Rapide-2.
Next, I compared the extent of the 170 000 prsim simulations light green font and also the mean green line. The observations forced into PRSIM (57 years) are in red. I plotted the maximum, the mean and the minimum.
I will show some of the 43 basins. As you can see, most of the means seem to be well simulated.
Gatineau
library(plotly) library(readr) library(tidyverse) # load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata') Bask<-as.data.frame(tests['Baskatong']) Qobs_prsim_mean<-Bask %>% group_by(Baskatong.MM,Baskatong.DD) %>%summarise(AvgQ=mean(Baskatong.Qobs,na.rm = TRUE)) Qobs_prsim_max<-Bask %>% group_by(Baskatong.MM,Baskatong.DD) %>%summarise(MaxQ=max(Baskatong.Qobs,na.rm = TRUE)) Qobs_prsim_min<-Bask %>% group_by(Baskatong.MM,Baskatong.DD) %>%summarise(MinQ=min(Baskatong.Qobs,na.rm = TRUE)) #Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019) path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/' fichiers<-list.files(path,pattern = '.csv') Q10000_etude2019<-read_csv(paste0(path,fichiers[1])) jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day") #Statistiques sommaires PRSIM load('/home/tito/Baskatong.Rdata') df_final_clean<-df_final[!duplicated(df_final$julian_day),] x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day") random_y <- df_final_clean$MaxQ data <- data.frame(x, random_y) trace1<-df_final_clean$AvgQ trace2<-df_final_clean$MinQ trace3<-Qobs_prsim_mean$AvgQ trace4<-Qobs_prsim_max$MaxQ trace5<-Qobs_prsim_min$MinQ fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% layout(title = "Average, High and Low PRSIM at Baskatong", paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)', xaxis = list(title = "Days", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE), yaxis = list(title = "Streamflow (cms)", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE)) fig
Outaouais Superieur
library(plotly) library(readr) library(tidyverse) # load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata') Bask<-as.data.frame(tests['Dozois']) Qobs_prsim_mean<-Bask %>% group_by(Dozois.MM,Dozois.DD) %>%summarise(AvgQ=mean(Dozois.Qobs,na.rm = TRUE)) Qobs_prsim_max<-Bask %>% group_by(Dozois.MM,Dozois.DD) %>%summarise(MaxQ=max(Dozois.Qobs,na.rm = TRUE)) Qobs_prsim_min<-Bask %>% group_by(Dozois.MM,Dozois.DD) %>%summarise(MinQ=min(Dozois.Qobs,na.rm = TRUE)) #Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019) path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/' fichiers<-list.files(path,pattern = '.csv') Q10000_etude2019<-read_csv(paste0(path,fichiers[1])) jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day") #Statistiques sommaires PRSIM load('/home/tito/Dozois.Rdata') df_final_clean<-df_final[!duplicated(df_final$julian_day),] x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day") random_y <- df_final_clean$MaxQ data <- data.frame(x, random_y) trace1<-df_final_clean$AvgQ trace2<-df_final_clean$MinQ trace3<-Qobs_prsim_mean$AvgQ trace4<-Qobs_prsim_max$MaxQ trace5<-Qobs_prsim_min$MinQ fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% layout(title = "Average, High and Low PRSIM at Dozois", paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)', xaxis = list(title = "Days", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE), yaxis = list(title = "Streamflow (cms)", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE)) fig
library(plotly) library(readr) library(tidyverse) # load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata') Bask<-as.data.frame(tests['Lac Victoria et lac Granet']) Qobs_prsim_mean<-Bask %>% group_by(Lac.Victoria.et.lac.Granet.MM,Lac.Victoria.et.lac.Granet.DD) %>%summarise(AvgQ=mean(Lac.Victoria.et.lac.Granet.Qobs,na.rm = TRUE)) Qobs_prsim_max<-Bask %>% group_by(Lac.Victoria.et.lac.Granet.MM,Lac.Victoria.et.lac.Granet.DD) %>%summarise(MaxQ=max(Lac.Victoria.et.lac.Granet.Qobs,na.rm = TRUE)) Qobs_prsim_min<-Bask %>% group_by(Lac.Victoria.et.lac.Granet.MM,Lac.Victoria.et.lac.Granet.DD) %>%summarise(MinQ=min(Lac.Victoria.et.lac.Granet.Qobs,na.rm = TRUE)) #Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019) path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/' fichiers<-list.files(path,pattern = '.csv') Q10000_etude2019<-read_csv(paste0(path,fichiers[1])) jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day") #Statistiques sommaires PRSIM load('/home/tito/Lac Victoria et lac Granet.Rdata') df_final_clean<-df_final[!duplicated(df_final$julian_day),] x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day") random_y <- df_final_clean$MaxQ data <- data.frame(x, random_y) trace1<-df_final_clean$AvgQ trace2<-df_final_clean$MinQ trace3<-Qobs_prsim_mean$AvgQ trace4<-Qobs_prsim_max$MaxQ trace5<-Qobs_prsim_min$MinQ fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% layout(title = "Average, High and Low PRSIM at Lac Victoria et lac Granet", paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)', xaxis = list(title = "Days", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE), yaxis = list(title = "Streamflow (cms)", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE)) fig
library(plotly) library(readr) library(tidyverse) # load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata') Bask<-as.data.frame(tests['Rapide-2']) Qobs_prsim_mean<-Bask %>% group_by(Rapide.2.MM,Rapide.2.DD) %>%summarise(AvgQ=mean(Rapide.2.Qobs,na.rm = TRUE)) Qobs_prsim_max<-Bask %>% group_by(Rapide.2.MM,Rapide.2.DD) %>%summarise(MaxQ=max(Rapide.2.Qobs,na.rm = TRUE)) Qobs_prsim_min<-Bask %>% group_by(Rapide.2.MM,Rapide.2.DD) %>%summarise(MinQ=min(Rapide.2.Qobs,na.rm = TRUE)) #Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019) path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/' fichiers<-list.files(path,pattern = '.csv') Q10000_etude2019<-read_csv(paste0(path,fichiers[1])) jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day") #Statistiques sommaires PRSIM load('/home/tito/Rapide-2.Rdata') df_final_clean<-df_final[!duplicated(df_final$julian_day),] x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day") random_y <- df_final_clean$MaxQ data <- data.frame(x, random_y) trace1<-df_final_clean$AvgQ trace2<-df_final_clean$MinQ trace3<-Qobs_prsim_mean$AvgQ trace4<-Qobs_prsim_max$MaxQ trace5<-Qobs_prsim_min$MinQ fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% layout(title = "Average, High and Low PRSIM at Rapide-2", paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)', xaxis = list(title = "Days", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE), yaxis = list(title = "Streamflow (cms)", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE)) fig
library(plotly) library(readr) library(tidyverse) # load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata') Bask<-as.data.frame(tests['Rapide-7']) Qobs_prsim_mean<-Bask %>% group_by(Rapide.7.MM,Rapide.7.DD) %>%summarise(AvgQ=mean(Rapide.7.Qobs,na.rm = TRUE)) Qobs_prsim_max<-Bask %>% group_by(Rapide.7.MM,Rapide.7.DD) %>%summarise(MaxQ=max(Rapide.7.Qobs,na.rm = TRUE)) Qobs_prsim_min<-Bask %>% group_by(Rapide.7.MM,Rapide.7.DD) %>%summarise(MinQ=min(Rapide.7.Qobs,na.rm = TRUE)) #Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019) path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/' fichiers<-list.files(path,pattern = '.csv') Q10000_etude2019<-read_csv(paste0(path,fichiers[1])) jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day") #Statistiques sommaires PRSIM load('/home/tito/Rapide-7.Rdata') df_final_clean<-df_final[!duplicated(df_final$julian_day),] x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day") random_y <- df_final_clean$MaxQ data <- data.frame(x, random_y) trace1<-df_final_clean$AvgQ trace2<-df_final_clean$MinQ trace3<-Qobs_prsim_mean$AvgQ trace4<-Qobs_prsim_max$MaxQ trace5<-Qobs_prsim_min$MinQ fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% layout(title = "Average, High and Low PRSIM at Rapide-7", paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)', xaxis = list(title = "Days", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE), yaxis = list(title = "Streamflow (cms)", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE)) fig
library(plotly) library(readr) library(tidyverse) # load('/media/tito/TIIGE/PRSIM/obs_outaouais.Rdata') Bask<-as.data.frame(tests['Lac des Quinze']) Qobs_prsim_mean<-Bask %>% group_by(Lac.des.Quinze.MM,Lac.des.Quinze.DD) %>%summarise(AvgQ=mean(Lac.des.Quinze.Qobs,na.rm = TRUE)) Qobs_prsim_max<-Bask %>% group_by(Lac.des.Quinze.MM,Lac.des.Quinze.DD) %>%summarise(MaxQ=max(Lac.des.Quinze.Qobs,na.rm = TRUE)) Qobs_prsim_min<-Bask %>% group_by(Lac.des.Quinze.MM,Lac.des.Quinze.DD) %>%summarise(MinQ=min(Lac.des.Quinze.Qobs,na.rm = TRUE)) #Scenario de Q10000 a Dozois dans l'etude de l'Outaouais Superieur (2019) path<-'/home/tito/Documents/github/prsim/outaouais_sup_lynda/Frequentielles/Apports_Lateraux_Methode_Extreme_prt/Q10000/' fichiers<-list.files(path,pattern = '.csv') Q10000_etude2019<-read_csv(paste0(path,fichiers[1])) jours_etude_2019<-seq(as.Date("2000/1/3"), as.Date("2000/1/3")+212, by = "day") #Statistiques sommaires PRSIM load('/home/tito/Lac des Quinze.Rdata') df_final_clean<-df_final[!duplicated(df_final$julian_day),] x <- seq(as.Date("2000/1/1"), as.Date("2000/12/31"), by = "day") random_y <- df_final_clean$MaxQ data <- data.frame(x, random_y) trace1<-df_final_clean$AvgQ trace2<-df_final_clean$MinQ trace3<-Qobs_prsim_mean$AvgQ trace4<-Qobs_prsim_max$MaxQ trace5<-Qobs_prsim_min$MinQ fig <- plot_ly(data, x = ~x, y = ~random_y, name = 'Qmax',type = 'scatter', mode = 'lines',line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace2, name = 'Qmin', mode = 'lines', type = 'scatter', fill = 'tonexty', fillcolor='rgba(0,100,80,0.2)', line = list(color = 'transparent'),showlegend = FALSE) fig <- fig %>% add_trace(y = ~trace1, name = 'Qavg', mode = 'lines',line = list(color='rgb(0,100,80)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace3, name = 'Qavg_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace4, name = 'Qmax_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% add_trace(y = ~trace5, name = 'Qmin_obs', mode = 'lines',line = list(color='rgb(255,0,0)'),type = 'scatter') fig <- fig %>% layout(title = "Average, High and Low PRSIM at Rapide-7", paper_bgcolor='rgb(255,255,255)', plot_bgcolor='rgb(229,229,229)', xaxis = list(title = "Days", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE), yaxis = list(title = "Streamflow (cms)", gridcolor = 'rgb(255,255,255)', showgrid = TRUE, showline = FALSE, showticklabels = TRUE, tickcolor = 'rgb(127,127,127)', ticks = 'outside', zeroline = FALSE)) fig
Future work
- Increase the probability of generation (Kappa)
- Do seasonality analysis (Boxplots per season as in Manuela's vignettes)
- Cross-correlation
- PACF, ACf (Replicate Manuela's vignettes with our 170 000 simulations)
- Madogram (Explore Pierre Ribatet package. Try to replicate P-Madogram in Manuela's article)
Questions
- What to do with Rapide-2 type of case?
- Could you please explain more in depth the interpretation of PACF, ACF and P-Madogram?
Thank you!
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.