data_analysis.Rmd

---
title: "TTD studies analysis"
output: 
  html_document:
    toc: true
---

```{r message=FALSE, warning=FALSE, include=FALSE}

# LOAD PACKAGES AND PROCESS THE TABLES

# load all packages 
library(readr)
library(dplyr)
library(tidyr)
library(stringr)
library(glue)
library(tibble)
library(knitr)
library(kableExtra)
library(DT)
library(ggplot2)
library(ggmap)
library(leaflet)
library(htmlwidgets)
library(htmltools)
```

This document explores the list of scientific studies included in the 'data' folder


```{r message=FALSE, warning=FALSE, include=FALSE}
# some settings
data_dir <- file.path("data") #directory where all the data is

# load some functions
source(file.path("scripts","functions.R"))

# process all the data through script "data_import.R" and import 3 tibbles:
  # T, with all the data on the studies and their sites (not really needed)
  # Tdisplay, which is the same as T but with html entries that are good for html display
  # Tsites, with all the info on the study sites
source(file.path("scripts","data_import.R"))

# some settings
file_export <- TRUE #to export figures
```

## TABLES

### The studies

The table below lists the considered `r nrow(Tdisplay)` applications, from `r length(unique(Tdisplay$study))` papers published between `r min(Tdisplay$year)` and `r max(Tdisplay$year)`.

```{r echo=FALSE, message=FALSE, warning=FALSE, include=TRUE}
# display as an html datatable
pglngth <- 10 #number of lines to display per page
table_studies <- generate_table_studies(Tdisplay,pglngth)
table_studies
```


### The study sites

The table below shows the main characteristics of the `r T %>% distinct(site, system) %>% count()` distinct sites where the studies were conducted.

```{r echo=FALSE, message=FALSE, warning=FALSE, include=TRUE}
# display as an html datatable
pglngth <- 10 #number of lines to display per page
table_sites <- generate_table_sites(Tsites,pglngth)
table_sites
```

## Map of the study sites

```{r echo=FALSE,  message=FALSE, warning=FALSE, fig.width=10, include = TRUE}
# display the html map
map_sites <- generate_map_sites(Tdisplay)
map_sites
# if (file_export){
#   saveWidget(map_sites,"Map.html", title="Map TTD sites") #save to a separate html file
#   file.rename("Map.html", file.path("html_files","Map.html"))
# }
```


## FIGURES

### TTD approaches

As important theoretical and experimental work consolidated between 2010 and 2015, the number of applications steadily grew in the following years.

```{r echo=FALSE, message=FALSE, warning=FALSE}
yrthresh <-  2006
F1 <- Tdisplay %>%
  filter(as.numeric(year) > yrthresh & as.numeric(year) < max(as.numeric(year))) %>% 
  distinct(study, .keep_all = TRUE) %>%
  arrange(year) %>% 
  ggplot(mapping = aes(x = as.numeric(year),
                       fill = `TTD approach`
                       ))+
  geom_bar()+
  labs(#title = "Publications 2006-2021, grouped by TTD approach",
       #subtitle = " ",
       #caption = "Figure S1",
       x = "year",
       y = "number") +
  theme_classic() 
F1

if (file_export){
  ggsave(file.path("figures","SI_approaches.pdf")) #save to separate pdf file
}
```


### Sites' size

```{r echo=FALSE, message=FALSE, warning=FALSE, fig.height = 3}
# save info on largest and smallest catchment
area_max <- Tsites %>% filter(system=="Catchment") %>% slice_max(size_km2)
area_min <- Tsites %>% filter(system=="Catchment") %>% slice_min(size_km2)
```

The catchments' areas range between `r area_max$size_km2` km^2 (`r area_max$site`, `r area_max$country`) and `r area_min$size_km2` km^2 (`r area_min$site`, `r area_min$country`). The median catchment size is `r median(as.numeric(Tsites$size_km2),na.rm = TRUE)` km^2.

```{r echo=FALSE, message=FALSE, warning=FALSE, fig.height = 3}
# make the figure with the site sizes
F2 <- Tsites %>%
  filter(system=="Catchment") %>% 
  dplyr::select(site,size_km2) %>% 
  distinct(site,size_km2) %>% 
  mutate(tmp = 1) %>% 
  mutate(randoffset = 0*size_km2+runif(sum(tmp), min = 0, max = 1)) %>% 
  ggplot(mapping = aes(x = size_km2,
                       y = "",
                       label = site))+
  geom_boxplot(color = "grey", alpha = .3)+
  geom_jitter(color = "red", alpha = .8, height = 0.2, width = 0)+
  #geom_point(color = "red", alpha = .8)+
  scale_x_continuous(trans='log10',
                     breaks = c(0.01, 0.1, 1, 10, 100),
                     label = c(0.01, 0.1, 1, 10, 100))+
  #ggrepel::geom_text_repel(size = 3, 
  #                         force = 1,
  #                         min.segment.length = 0) +
  labs(#title = "Catchments' size",
       #subtitle = " ",
       #caption = "Figure S2",
       x = "Size (km2)",
       y = "") +
  theme_classic()
F2

if (file_export){
  ggsave(file.path("figures","SI_sites_size.pdf"),width = 8, height = 3)  #save to separate pdf file
}
```


### Climates of the study sites

The large majority of the studies were carried out in temperate and continental/cold climates ('C' and 'D' main codes, respectively), mainly without a dry season. Tropical and dry climates ('A' and 'B') are almost absent from the list. This reflects the known bias of hydrologic research for experimental sites located at mid-latitudes in Europe and the United States. Climate classification (after [Beck et al., 2018](https://doi.org/10.1038/sdata.2018.214)) is based on one single point and may thus be inaccurate for large catchments.

```{r echo=FALSE,  message=FALSE, warning=FALSE}
  # show the climates
F3 <- Tdisplay %>%
  #filter(system==c("Catchment")) %>% 
  filter(!is.na(climate_id)) %>% 
  distinct(site, .keep_all = TRUE) %>%
  mutate(climate_code_main = str_sub(climate_code,1,1),
         climate_description = str_c(climate_code,climate_description, sep = ": ")) %>% 
  ggplot(mapping = aes(x = climate_code_main,
                       fill = climate_description))+
  geom_bar(color = "white")+
  #coord_flip() +
  labs(title = "Sites by Köppen-Geiger climate",
       subtitle = "Classification according to Beck et al. (2018)",
       #caption = "Figure S3",
       x = "class",
       y = "count",
       fill = "Climate codes") +
  theme_classic() 
F3

if (file_export){
  ggsave(file.path("figures","SI_sites_climate.pdf")) #save to separate pdf file
}
```