Creating near real-time global 10-m land cover maps with geemap and Dynamic World

App: https://www.dynamicworld.app
App2: https://earthoutreach.users.earthengine.app/view/dynamicworld
Paper: https://doi.org/10.1038/s41597-022-01307-4
Model: https://github.com/google/dynamicworld
Training data: https://doi.pangaea.de/10.1594/PANGAEA.933475
Data: https://developers.google.com/earth-engine/datasets/catalog/GOOGLE_DYNAMICWORLD_V1
JavaScript tutorial: https://developers.google.com/earth-engine/tutorials/community/introduction-to-dynamic-world-pt-1

In [ ]:

import ee
import geemap

In [ ]:

Map = geemap.Map()
Map.add_basemap('HYBRID')
Map

In [ ]:

# Set the region of interest by simply drawing a polygon on the map
region = Map.user_roi
if region is None:
    region = ee.Geometry.BBox(-89.7088, 42.9006, -89.0647, 43.2167)

Map.centerObject(region)

In [ ]:

# Set the date range
start_date = '2021-01-01'
end_date = '2022-01-01'

In [ ]:

# Create a Sentinel-2 image composite
image = geemap.dynamic_world_s2(region, start_date, end_date)
vis_params = {'bands': ['B4', 'B3', 'B2'], 'min': 0, 'max': 3000}
Map.addLayer(image, vis_params, 'Sentinel-2 image')

In [ ]:

# Create Dynamic World land cover composite
landcover = geemap.dynamic_world(region, start_date, end_date, return_type='hillshade')
Map.addLayer(landcover, {}, 'Land Cover')

In [ ]:

# Add legend to the map
Map.add_legend(title="Dynamic World Land Cover", builtin_legend='Dynamic_World')
Map

In [ ]:

# Save Dynamic World class data in GeoTIFF format
output_path = 'landcover.tif'
landcover = geemap.dynamic_world(region, start_date, end_date, return_type='class')
geemap.ee_export_image(landcover, filename=output_path, scale=10, region=region, file_per_band=False)