Exercise1_sdmpredictors.Rmd

---
title: 'Exercise 1 - sdmpredictors'
output:
  html_document:
    toc: true
    toc_depth: 5
    toc_float: true
    df_print: paged
---

The purpose of this exercise is to check the maximum sea surface temparature near


* Key West (24.545°N, 81.775°W)
* Barbados (13.185° N, -59.440° W)
* Porto (41.130° N, -8.882° W)?

***

<br>


#### 1.1. Data: sdmpredictors

Where can we look for the data?
<http://www.bio-oracle.org/>

##### 1.1.2 prepare R environment
First, we load some R packages that we will use in this exercise
```
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
library(sf) # spatial package
library(ggplot2) # plotting
library(sdmpredictors) # species distribution modeling layers => data layers
```

Now, we want to use the data from the sdmpredictors package

##### 1.1.2 Look for correct data layer

sdmpredictors contains several datasets, each containing several layers:
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# list datasets of sdmpredictors package
list_datasets()

# list layers of Bio-ORACLE dataset
list_layers(datasets="Bio-ORACLE")
```

##### 1.1.3 Download layer

Now we store the data from the layer to R:
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
sstmax <- load_layers("BO_sstmax")

```

***

<br>

#### 1.2 Plotting data

##### 1.2.1 Base R Plot

Let's try R's base plot:
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# base plot
plot(sstmax)
```

This is not so user friendly, let's try to make an interactive map.

##### 1.2.2 Mapview interactive map

With the 'mapview' package, you can easily create interactive maps.
For more info, see <https://r-spatial.github.io/mapview/>

```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# Make an interactive plot with the 'mapview' package:
# more info: https://r-spatial.github.io/mapview/
library('mapview')
```

if we just try `mapview(sstmax)`, there is an error. What does it say?

```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
mapview(sstmax, layer.name = "max sst") # warning: what does it say?


mapview(sstmax, layer.name = "max sst", maxpixels =  9331200) # takes a while to load!

```

***

<br>


#### 1.3 Points of interest
##### 1.3.1 create dataframe

Let's create a dataframe with our points of interest:

```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
my.sites <- data.frame(
  Name = c("Key West", "Barbados", "Porto"),
  Lon = c(-81.775,-59.440,-8.882),
  Lat = c(24.545,13.185,41.130))
```
##### 1.3.2 make spatial 'simple features' 

```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# make spatial (simple features) dataset
my.sites.sf <- st_as_sf(my.sites,
                        coords=c("Lon","Lat"),
                        crs = 4326) # WGS84
```

##### 1.3.3 base plot locations
Again, R's base plotting capabilities are not ideal...
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# base plot
plot(my.sites.sf) # not ideal
```
##### 1.3.3 interactive mapview map 
An interactive map is a lot better, right?
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# interactive mapview map
mapview(my.sites.sf) # much better
```

##### 1.3.3 visualizing both layers
How do we visualize both of the layers?
Just combine them with a plus sign: '+':
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# how to visualize both? combine them with '+'
mapview(sstmax, layer.name = "max sst") +
  mapview(my.sites.sf)
```

***

<br>


#### 1.4 extracting data
Now, we want to get the sea surface temperature data at the exact location of our points.
This is easily done with the `raster` package function `extract`:
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# extract data from sstmax raster at the locations of my.sites.sf
my.sites$BO_sstmax <- extract(sstmax,my.sites.sf)
```

***

<br>


#### 1.5 Plotting data in ggplot


We will plot the data that we have with the *ggplot2* package:
<https://ggplot2.tidyverse.org/>

The grammar of graphics say that can build any graph with the same components:

* the data(set)
* aesthetics: how you position your data
* a geometry: how to visualize the data points
* (some additional styling)


```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# plotting data with ggplot2
# https://ggplot2.tidyverse.org/
library(ggplot2)
```

Let's create an easy example with

* data: `my.sites`
* aesthetics: along x-axis the `Name` column, y-axis `BO_sstmax` (the temperature)
* geometry: points

```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# points
ggplot(data = my.sites) +
  geom_point(aes(x = Name, y = BO_sstmax))
```


If we improve the graph by:

*  color filling the points by the sea surface temperature value
*  use the same color scale as the sst map
*  make the pionts a bit bigger

```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
  # with different colors, a bit larger size
  ggplot(data = my.sites) +
    geom_point(aes(x = Name,
                   y = BO_sstmax,
                   color = BO_sstmax,
                   size = 2)) +
    scale_color_viridis_c(option = 'inferno') # same colorscale as map
```

We can change the geometry to a barplot:
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
# bar plot
ggplot(data = my.sites) +
  geom_col(aes(x = Name, y = BO_sstmax))
```

different colors, larger size (outline width):
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
  # again different colors, larger size:
  ggplot(data = my.sites) +
    geom_col(aes(x = Name,        # everything inside aes(), will be 'interpreted'
                 y = BO_sstmax,
                 fill = Name),
                 size = 2,
                 color = "black") # color : outline with graphs
```
Some more adjustments
```{r message=FALSE, warning=FALSE, error=FALSE, fig.align='center'}
  # color by sst max
  ggplot(data = my.sites) +
    geom_col(aes(x = Name,        # everything inside aes(), will be 'interpreted'
                 y = BO_sstmax,
                 fill = BO_sstmax),
             color = "black") +
             scale_fill_viridis_c(option = 'inferno') # color : outline with graphs
```

***

<br>


#### Exercise 1
Adjust the code above to include 3 areas of your interest

e.g. the location closest you your lab?

#### Advanced exercise 1
What is the temperature at different latitudes in the North Atlantic?

e.g. at longitude -30, latitude from 0 to 60?
 

#### Advanced exercise 2
What is the minimum temperature?
Can you plot:

* the min temperature vs max temperature
* min AND max temp in a plot?