In BigelowLab/ohwobpg: ohwobpg for Ocean Biology Processing Group data

knitr::opts_chunk$set(echo = TRUE)

Install and load packages

# Load libraries
library(raster)
library(ggplot2)
library(leaflet)
library(dplyr)
library(remotes)
# Install ohwobpg package
installed <- rownames(installed.packages())
if (!("ohwobpg" %in% installed))  remotes::install_github("BigelowLab/ohwobpg", quiet = TRUE)
# Load ohwobpg package
library(ohwobpg)

Initialize raster for visualization

Once the necessary packages are loaded, we will start by creating a raster object from a pre-loaded ohwobpg database. Specifically, we will be creating a raster object of sea surface temperature (SST) data for June of 2018.

# Load database
path <- system.file("gom", package = "ohwobpg")
db <- read_database(path)

# Subset database
par_db <- db %>% 
  dplyr::filter(param == "sst" & 
                per == "MO" &
                dplyr::between(date, as.Date("2018-05-15"), as.Date("2018-09-26")))

# Create raster stack
par <- par_db %>%                # start with the subset database
  as_filename(path = path) %>%   # build filenames and append to the path
  raster::stack()                # read them into a stack of images

# Name layers
names(par) <- format(par_db$date, "%b")

# Isolate June
r <- par$Jun

Plot rasters using base plot

Next, we will plot the June 2018 SST raster object using the base R plot() function. We simply call the plot() function with the raster object as the argument. For more options to make fancier plots with base plot and a more detailed example, see https://www.neonscience.org/dc-plot-raster-data-r.

plot(r)

Plot rasters using ggplot2

Next, we will plot the same raster object using the package ggplot2. First, the user needs to convert the raster object to a dataframe with coordinates, using as.data.frame(raster, xy = TRUE). Next, the dataframe can be plotted using geom_tile(). For an in-depth example of how to plot a raster using ggplot2 and geom_tile(), see https://ggplot2.tidyverse.org/reference/geom_tile.html.

# Convert raster to dataframe
r_df <- as.data.frame(r, xy = TRUE)
# Plot dataframe using ggplot2
ggplot(data = r_df, mapping = aes(x = x, y = y, color = Jun, fill = Jun)) +
  # Add raster data
  geom_tile() +
  # Change NA fill color to white
  scale_fill_continuous(na.value = "white") +
  scale_color_continuous(na.value = "white") +
  # Add axes labels
  labs(x = "Lon",
       y = "Lat")

Plot rasters using leaflet

Finally, we will plot the raster object on a leaflet interactive map. The leaflet::addRasterImage() function allows the user to plot raster objects on the map. For a more detailed example, see https://rstudio.github.io/leaflet/raster.html.

leaflet::leaflet() %>% leaflet::addTiles(group = "Standard") %>% 
            # Add satellite imagery
            leaflet::addProviderTiles('Esri.WorldImagery', group = "Satellite") %>%
            # Define bounds
            leaflet::fitBounds(lng1 = min(r_df$x), 
                               lat1 = min(r_df$y), 
                               lng2 = max(r_df$x), 
                               lat2 = max(r_df$y)) %>%
            # Add raster data
            leaflet::addRasterImage(r)