Styjun

I have a high-resolution coastline shapefile from NOAA that covers a large chunk of the globe. My aim is to:

1 - Clip this dataset to a smaller area (already achieved OK);
2 - Overlay the bounded area on a base R plot (still having problems).

Here's my code:

library(rgeos)
library(rgdal)

download.file(url = 'https://dnc.nga.mil/DNCSiteContent/Shapefiles/shapefile01.zip', destfile = '~/Desktop/shapefile01.zip')
unzip('~/Desktop/shapefile01.zip', exdir = '~/Desktop/shapefile01')
closeAllConnections()
system('rm -rf ~/Desktop/shapefile01.zip')
targetfolder <- dir('~/Desktop', pattern = 'shapefile01', full.names = TRUE) # need full directory path for rgdal::readOGR

shape <- rgdal::readOGR(dsn = targetfolder, layer = 'cd01')
bbx <- rgeos::readWKT('POLYGON((-42.05 -23.03, -41.95 -23.03, -41.95 -22.93, -42.05 -22.93, -42.05 -23.03))')
proj4string(bbx) <- proj4string(shape)
mysite <- rgeos::gIntersection(shape, bbx)
plot(mysite)

What I ideally want here is to color the terrain on the map, such that it would look like this (here edited with Photoshop):

But mysite, an object of class SpatialLines with 28 line segments that should be a single thing, but in reality are not connected. So when I plot the whole object, the line contour of this coastal area appears to be just fine. But because they're a bunch of smaller line segments, R doesn't know how to put them together so I can fill the terrain. Is there a way to tell R to color fill the continent? Maybe combine all these segments into one continuous continental contour?

By the way, I would ideally like to do this in base R because this will be later added (i.e. overlaid) onto another plot with a color gradient fill for the ocean around the same geographical area (generated from another spatial file).

UPDATE

Using the spatstat package I can now visualise all the line segments very easily, so I thought it would be nice to update this here to help visualise the problem. Here's the code:

library(spatstat)
mysite2 <- spatstat::as.psp(mysite)
plot(mysite2)

Makes sense? I think it would be nice to have a permanent solution because that's the nature of these high-res shoreline shapefiles provided by NOAA.