05-polishing-your-map.Rmd

# Polishing Your Map for Publication

## Introduction

In this section we will polish the map for publication.

Maps traditionally have several specific elements that help them convey the information to the user.
These include legends, which we've already added to our maps, north arrows to orient the user and scale bars and scale to show the scale of the map.

Projections are used to represent the spherical surface of the Earth in a flat plane. 
Think of wrapping a piece of paper around a ball.
It does not wrap evenly, a map is the reverse process, trying to flatten a globe or sphere.
Projections help us deal with these distortions.
Published maps should always be projected for the area of interest.
As the area gets smaller, the projection becomes less important.
The data that we downloaded from Naturalearthdata and Worldclim both are in an [Equirectangular projection](https://en.wikipedia.org/wiki/Equirectangular_projection).

In the exercises, first, we will re-project the data to SIRGAS 2000, which is what the Brazilian government normally uses for their maps.
Then we will add elements to complete our map including the north arrow, scale bar and text, title, data attribution and a neatline.

## Exercises

### Re-project Data

There are several types of map projection depending on the scale (_e.g._ global, continent, national, state).
Continental and national scale projections often depend on the location on the globe to help inform the choice of projection to use.

First, we check the current projections of the data.

```{r re-project}
library(sf)
st_crs(br_sf)
```

To reproject the data we will provide the definition for SIRGAS 2000.
The information for the `proj4text` comes from <https://epsg.io/4674>, which has information for several different map projections.

```{r sirgas-2000}
br_sf <- st_transform(br_sf, 4674)
st_crs(br_sf)
```

### Adding North Arrows and Scale Bars

Using _ggspatial_ we will add a scale bar and north arrow to the map of Minas Gerais to the bottom right corner where there is a bit of space for them.

```{r n-arrow, message=FALSE}
library("ggspatial")

ggplot() +
  geom_sf(data = br_sf) +
  geom_raster(data = brazil, aes(x = x, y = y, fill = tmin1)) +
  scale_fill_viridis_c("Min\nT (°C)", na.value = NA) +
  xlab("Longitude") +
  ylab("Latitude") +
  annotation_north_arrow(
    location = "br",
    which_north = "true",
    pad_x = unit(0.25, "in"),
    pad_y = unit(0.25, "in"),
    style = north_arrow_fancy_orienteering
  ) +
  coord_sf() +
  theme_bw()
```

Next we will add the scale bar to the bottom right where there is a bit more space and it won't interfere with the map.

```{r scale-bar-1, message=FALSE}
br_tmin <- ggplot() +
  geom_sf(data = br_sf) +
  geom_raster(data = brazil, aes(x = x, y = y, fill = tmin1)) +
  scale_fill_viridis_c("Min\nT (°C)", na.value = NA) +
  xlab("Longitude") +
  ylab("Latitude") +
  annotation_north_arrow(
    location = "br",
    which_north = "true",
    pad_x = unit(0.25, "in"),
    pad_y = unit(0.25, "in"),
    style = north_arrow_fancy_orienteering
  ) +
  annotation_scale(location = "br", width_hint = 0.5) +
  coord_sf() +
  theme_bw()
```

### Adding a Descriptive Title and Data Source

Adding a title is the same as any other _ggplot2_ object.

```{r map-title, message=FALSE}
br_tmin +
  labs(title = "Map of Minimum Average Temperature")
```

It is always a good idea to credit the sources of the map data, especially if it is not your own.
The `labs()` has options for subtitles, which can be useful, but also a `caption` argument, which is useful for adding data attributions to your map.

```{r data-attribution, warning=FALSE, message=FALSE}
br_tmin <-
  br_tmin +
  labs(title = "Map of Minimum Average Temperature",
       subtitle = "Month of January",
       caption = "Data from Worlclim, used with permission")

br_tmin
```

Neatlines are traditionally part of a map but in many cases nowadays, it is not uncommon to see maps without them.
If you wish to add a neatline to your map it is possible to do using the _gridExtra_ and _grid_ packages.

```{r neatline}
library(grid)
library(gridExtra)

grid.arrange(br_tmin, ncol = 1)
grid.rect(
  width = 0.98,
  height = 0.98,
  gp = gpar(lwd = 0.25,
            col = "black",
            fill = NA)
)
```

## Your Turn {.exercise}

* Using the examples above, create a map of a state in Brazil of your choice, add all of the map elements from above to make a complete map of just the state.

* Using the datavfor the map of tmin for Minas Gerais you created in the last exercise, create a map of tmin for Minas Gerais and add all the appropriate map elements to it.