This data derived from Natural Earth 1 portrays the world environment in an idealized manner with little human influence. Natural Earth I is available with ocean bottom data, or without. Satellite-derived land cover data and shaded relief presented with a light, natural palette suitable for making thematic and reference maps. As in nature, the map colors gradually blend into one another across regions (x and y axis) and from lowlands to highlands (z axis), hence the name cross-blended hypsometric tints. Shaded relief combined with custom elevation colors based on climate-humid lowlands are green and arid lowlands brown. Note: Choose a category below to see matching downloads.įiles have been downloaded 903,764 times. Embedded raster content includes: land cover, shaded relief, ocean water, and drainages with lakes. The raster files register precisely with the 10m vector data. Two versions of the 10 million-scale raster data are offered: high resolution files at 21,600 x 10,800 pixels and low resolution at 16,200 x 8,100. theme_map <- function(.Natural Earth features 7 types of raster files at 1:10 million-scale to suit your bandwidth and content focus. However, you can set these to negative values to remove the “excess” border around plots should you want to have a tight crop. We then apply this theme to the previous plot.įor clarity the margins are kept at 0 which generally does not influence general plotting. It sets all axis, tick marks, and grids to empty ( element_blank()), while specifying the background colour of the map (lightblue). The theme below specifies the font used, and colour of the text. You can tackle these map wide features by either adjusting a theme inline or creating a separate theme function to append to your ggplot2 layout. Seed = 1 # ensures the placing is consistent between rendersĪt this point we have a rather clean map but you can argue that given the size of the map we can do away with the grid lines and axis labels. The end points of the label segments (lines) are accentuated with a simple grey point rendered by geom_point(). When this value is random, your placement of labels will be so too. This is set to a fixed value in order to maintain the label positions stable between renders. This makes it possible for you not to worry about the placement of the labels as these are assigned automatically in order to avoid overlapping. Note that we use the ggrepel package and the geom_text_repel() function for this. This is required to plot the country boundaries over the raster data. Note that we’ll add the country layer again, but without a fill value. We’ll add the raster data with forest locations first, giving it a dark olive green colour. This is a cheeky workaround for this issue. However, with a large area in a non-planar projection this is not possible. When dealing with planar projections you can use the sf function st_buffer() to calculate a true buffer around your polygons. Together with the land mass boundaries plotted before this generated a faux drop shadow effect. In the next step we add the countries, however we use a slightly thinner line width for the outlines of the countries. Normally, you do this at the end of your routine, but to keep the maps small I’ll repeat this part in each step. This is a trick which will become obvious in the next step. We’ll first lay down the basic outline of the land area, this is the foundation of our map pie. For illustration purposes we’ll cover this step by step, and showing the intermediate maps. Now we have all the data we can start building the final plot. # read vector data using rnaturalearthįilter(lc != 0) # only retain pixels with a true content We also generate some fake points of interest. We also read in a layer of MODIS land cover classes, which can be downloaded from the LP DAAC. So we’ll use the rnaturalearth functions to download both outlines of the land surface, and a map of countries. # read librariesĪfter loading all these packages we need to access the data we want to use in this project. Finally, both ggrepel and showtext are there to provide fancy formatting for text. In addition, we’ll need the rnaturalearth package to download vector based map data. These include general data wrangling packages tidyverse, and the sf and raster packages to deal with vector and raster based geo-spatial data. Before we begin building the map we must load some required libraries to assist in our map building work.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |