Creating a ggplot2 Map of the US Engineering Weather Sites

install.load::load_package("ie2miscdata", "USA.state.boundaries", "data.table", "spatstat", "sp", "rgdal", "ggplot2", "spatstat.geom", "maptools")
# load needed packages using the load_package function from the install.load package (it is assumed that you have already installed these packages)


data(weather_results) # load the weather_results data (containing the site information for US weather stations)
data(USA_state_boundaries_map) # load the USA_state_boundaries_map data (for the US map)

weather_results_map <- copy(weather_results) # copy the weather_results using data.table
setnames(weather_results_map, 3:4, c("lat", "lon")) # set the names of columns 3 and 4 using data.table


# Source 1 - 6 begins
# set coordinates and projection
class(weather_results_map)
## [1] "data.table" "data.frame"
coordinates(weather_results_map) <- ~ lon + lat
class(weather_results_map)
## [1] "SpatialPointsDataFrame"
## attr(,"package")
## [1] "sp"
proj4string(weather_results_map) <- CRS("+proj=longlat +datum=NAD83")

weather_results_map <- spTransform(weather_results_map, CRS(proj4string(USA_state_boundaries_map)))

weather_results_map <- data.frame(weather_results_map)


# Switch to x/y from lat/lon
# change variable names
names(weather_results_map)[names(weather_results_map) == "lon"] <- "x"
names(weather_results_map)[names(weather_results_map) == "lat"] <- "y"


# read in shapefile with rgdal
US <- USA_state_boundaries_map
USregions <- slot(US, "polygons")
USregions <- lapply(USregions, function(x) { SpatialPolygons(list(x)) })
USwindows <- lapply(USregions, as.owin)
USh <- hyperframe(window = USwindows)
USh <- cbind.hyperframe(USh, US@data)

USowin <- as(US, "owin") # Source 7


# determine which locations are within the borders of the US (including Alaska, Hawai'i, and Puerto Rico)
insideUS <- which(inside.owin(weather_results_map$x, weather_results_map$y, USowin)) # Source 7

weather_results_keep <- weather_results[insideUS, ]

line1 <- "Design wet bulb temperatures - 1% exceedance values"

# create a data.table copy of the sites that are kept
weather_results_map_keep <- copy(weather_results_keep)
setnames(weather_results_map_keep, 3:4, c("lat", "lon"))


# set coordinates and projection for the locations contained within the US
class(weather_results_map_keep)
## [1] "data.table" "data.frame"
coordinates(weather_results_map_keep) <- ~ lon + lat
class(weather_results_map_keep)
## [1] "SpatialPointsDataFrame"
## attr(,"package")
## [1] "sp"
proj4string(weather_results_map_keep) <- CRS("+proj=longlat +datum=NAD83")

weather_results_map_keep <- spTransform(weather_results_map_keep, CRS(proj4string(USA_state_boundaries_map)))

weather_results_map_keep <- data.frame(weather_results_map_keep)


# Switch to x/y from lat/lon
# change variable names
names(weather_results_map_keep)[names(weather_results_map_keep) == "lon"] <- "x"
names(weather_results_map_keep)[names(weather_results_map_keep) == "lat"] <- "y"


# plot the map using ggplot2
colorsnow <- c("Weather Data Sites" = "#3591d1")

p <- ggplot() + geom_polygon(data = USA_state_boundaries_map, aes(x = long, y = lat, group = group), colour = "black", fill = "white")
p <- p + geom_point(data = weather_results_map_keep, aes(x = x, y = y, color = "Weather Data Sites"))
p <- p + labs(x = "", y = "", title = "US Engineering Weather Sites Map")
p <- p + scale_colour_manual(values = colorsnow, labels = c("Weather Data Sites"), name = "Legend")
p <- p + theme(axis.ticks.y = element_blank(), axis.text.y = element_blank(), # get rid of x ticks/text
          axis.ticks.x = element_blank(), axis.text.x = element_blank(), # get rid of y ticks/text
          plot.title = element_text(lineheight = 0.8, face = "bold", vjust = 1), # make title bold and add space
          panel.grid.major = element_blank(), # Source 3 / get rid of major grid
          panel.grid.minor = element_blank()) # Source 3 / get rid of minor grid
print(p)

# Source 1 - 6 ends



Sources

Mapping in R using the ggplot2 package Posted on July 16, 2014 by ZevRoss. See http://zevross.com/blog/2014/07/16/mapping-in-r-using-the-ggplot2-package/.


use proj4 to specify Robinson projection with R ggmap and ggplot2 packages? - Geographic Information Systems Stack Exchange asked by Abe on Dec 19 2012 & answered by radek on Dec 20 2012. See https://gis.stackexchange.com/questions/44387/use-proj4-to-specify-robinson-projection-with-r-ggmap-and-ggplot2-packages.


r - How to increase size of the points in ggplot2, similar to cex in base plots? - Stack Overflow answered by Gavin Simpson on Mar 20 2012. See https://stackoverflow.com/questions/9789613/how-to-increase-size-of-the-points-in-ggplot2-similar-to-cex-in-base-plots.


r - ggplot2: add color to boxplot - "Continuous value supplied to discrete scale" error - Stack Overflow answered by Brian Diggs on May 29 2012. See https://stackoverflow.com/questions/10805643/ggplot2-add-color-to-boxplot-continuous-value-supplied-to-discrete-scale-er.


r - Interval size and colors for points on USA map - discrete error - Stack Overflow answered and edited by zx8754 on Mar 23 2015. See https://stackoverflow.com/questions/29216265/interval-size-and-colors-for-points-on-usa-map-discrete-error.


r - remove grid, background color and top and right borders from ggplot2 - Stack Overflow edited by Sandy Muspratt on Apr 17 2015. See https://stackoverflow.com/questions/10861773/remove-grid-background-color-and-top-and-right-borders-from-ggplot2.


R-sig-geo - Problem in converting SpatialPolygonsDataFrame to owin object Answer by Roger Bivand on Sep 15, 2006. See https://stat.ethz.ch/pipermail/r-sig-geo/2006-September/001313.html.