global-land-mask
Check whether a lat/lon point is on land for any point on earth.
Install
Install using pip:
pip install global-land-mask
Requirements:
Description
This python module, global-land-mask, contains scripts for checking whether a lat/lon point is on land or sea. In order to do this, we use the GLOBE dataset, which samples the entire earth at 1 km resolution. We then simply extract all 'invalid' values from this elevation map and save to file.
The global mask is of shape (21600, 43200), equating to about 980 mB when saved without compression. This data can be compressed to 2.5 mb using numpy savez_compressed, making for a very compact package.
The raw elevation data from the GLOBE dataset can be downloaded from
https://www.ngdc.noaa.gov/mgg/topo/gltiles.html
It is not necessary to download this data in order to use the global land mask. However, the user can build a global elevation dataset using the exampel scripts provided.
This package provides globe.is_land(), an alaternative to Basemap.is_land(). For 10,000 data points, globe.is_land is around 6000 times faster than Basemap.is_land.
The function globe.is_land uses a nearest-neighbor lookup on a regular grid.
Simple example
Here is a simple example showing the use of global_land_mask to check if lat/lon points are on land.
from global_land_mask import globe
import numpy as np
lat = 40
lon = -120
is_on_land = globe.is_land(lat, lon)
print('lat={}, lon={} is on land: {}'.format(lat,lon,is_on_land))
lat = 40
lon = np.linspace(-150,-110,3)
is_in_ocean = globe.is_ocean(lat, lon)
print('lat={}, lon={} is in ocean: {}'.format(lat,lon,is_in_ocean))
Speed test
Compare performance of global_land_mask and Basemap.
from global_land_mask import globe
from mpl_toolkits.basemap import Basemap
import matplotlib
matplotlib.use('TkAgg')
import numpy as np
import time
lat = np.linspace(-20,50,100)
lon = np.linspace(-130,-70,100)
lon_grid, lat_grid = np.meshgrid(lon,lat)
start_time = time.time()
globe_land_mask = globe.is_land(lat_grid, lon_grid)
globe_run_time = time.time()-start_time
print('Time to run globe.is_land(): {}'.format(globe_run_time))
bm = Basemap(projection='cyl', llcrnrlat=-60, urcrnrlat=90, \
llcrnrlon=-180, urcrnrlon=180, resolution='c')
f = np.vectorize(bm.is_land)
start_time = time.time()
xpt, ypt = bm( lon_grid, lat_grid)
basemap_land_mask = f(xpt,ypt)
basemap_run_time = time.time()-start_time
print('Time to run Basemap.is_land(): {}'.format(basemap_run_time))
print('Speed up: {}'.format(basemap_run_time/globe_run_time))
fraction_agreed = np.sum(globe_land_mask == basemap_land_mask)/len(globe_land_mask.flatten())
print('Fraction agreeing: {}'.format(fraction_agreed))
Example of map over US
Try running
from global_land_mask import globe
from mpl_toolkits.basemap import Basemap
import numpy as np
import matplotlib
matplotlib.use('TkAgg')
import matplotlib.pyplot as plt
lat = np.linspace(-20,90,1000)
lon = np.linspace(-130,-60,1002)
lon_grid, lat_grid = np.meshgrid(lon,lat)
z = globe.is_land(lat_grid, lon_grid)
fig = plt.figure(0, figsize=(5.5, 4.5))
plt.clf()
ax = fig.add_axes([0.1, 0.1, 0.8, 0.8])
m = Basemap(llcrnrlon=-119, llcrnrlat=22, urcrnrlon=-64, urcrnrlat=49,
projection='lcc', lat_1=33, lat_2=45, lon_0=-95,
area_thresh=200,
resolution='i')
m.drawstates(linewidth=0.2)
m.drawcoastlines(linewidth=0.2)
m.drawcountries(linewidth=0.2)
cs = m.contourf(lon_grid, lat_grid, z,
levels=[-0.5, 0.5,1.5],
cmap="jet",
latlon=True)
plt.show()
plt.savefig('example_plot_globe_map_us.png',
bbox_inches='tight',
dpi=400)
to create the binary mask for the US, shown in the image below:
Note that lakes are included as "on land" and the resolution isn't perfect, but it's good enough for many purposes!