AGU 2023
An Introduction to Cloud-Based Geospatial Analysis with Earth Engine and Geemap
- Notebook: https://geemap.org/workshops/AGU_2023
- Earth Engine: https://earthengine.google.com
- Geemap: https://geemap.org
Introduction (10 mins)¶
This notebook is for the workshop presented at the AGU Annual Meeting 2023.
Abstract¶
This workshop provides an introduction to cloud-based geospatial analysis using the Earth Engine Python API. Attendees will learn the basics of Earth Engine data types and how to visualize, analyze, and export Earth Engine data in a Jupyter environment with geemap. In addition, attendees will learn how to develop and deploy interactive Earth Engine web apps with Python. Through practical examples and hands-on exercises, attendees will enhance their learning experience. During each hands-on session, attendees will walk through Jupyter Notebook examples on Google Colab with the instructors. At the end of each session, they will complete a hands-on exercise to apply the knowledge they have learned.
Prerequisites¶
To use geemap and the Earth Engine Python API, you must register for an Earth Engine account and follow the instructions here to create a Cloud Project. Earth Engine is free for noncommercial and research use. To test whether you can use authenticate the Earth Engine Python API, please run this notebook on Google Colab.
It is recommended that attendees have a basic understanding of Python and Jupyter Notebook.
Familiarity with the Earth Engine JavaScript API is not required but will be helpful.
Attendees can use Google Colab to follow this workshop without installing anything on their computer.
Introduction to Earth Engine and geemap (15 mins)¶
Earth Engine is free for noncommercial and research use. For more than a decade, Earth Engine has enabled planetary-scale Earth data science and analysis by nonprofit organizations, research scientists, and other impact users.
With the launch of Earth Engine for commercial use, commercial customers will be charged for Earth Engine services. However, Earth Engine will remain free of charge for noncommercial use and research projects. Nonprofit organizations, academic institutions, educators, news media, Indigenous governments, and government researchers are eligible to use Earth Engine free of charge, just as they have done for over a decade.
The geemap Python package is built upon the Earth Engine Python API and open-source mapping libraries. It allows Earth Engine users to interactively manipulate, analyze, and visualize geospatial big data in a Jupyter environment. Since its creation in April 2020, geemap has received over 2,900 GitHub stars and is being used by over 1,000 projects on GitHub.
Google Colab and Earth Engine Python API authentication (5 mins)¶
Change Colab dark theme¶
Currently, ipywidgets does not work well with Colab dark theme. Some of the geemap widgets may not display properly in Colab dark theme.It is recommended that you change Colab to the light theme.
Install geemap¶
The geemap package is pre-installed in Google Colab and is updated to the latest minor or major release every few weeks. Some optional dependencies of geemap being used by this notebook are not pre-installed in Colab. Uncomment the following line to install geemap and some optional dependencies.
# %pip install -U "geemap[workshop]"
Note that some geemap features do not work properly with Google Colab. If you are familiar with Anaconda or Miniconda, it is recommended to create a new conda environment to install geemap and its optional dependencies on your local computer.
conda create -n gee python=3.11
conda activate gee
conda install -c conda-forge mamba
mamba install -c conda-forge geemap pygis
Import libraries¶
Import the earthengine-api and geemap.
import ee
import geemap
Authenticate and initialize Earth Engine¶
You will need to create a Google Cloud Project and enable the Earth Engine API for the project. You can find detailed instructions here.
geemap.ee_initialize()
Creating interactive maps¶
Let's create an interactive map using the ipyleaflet
plotting backend. The geemap.Map
class inherits the ipyleaflet.Map
class. Therefore, you can use the same syntax to create an interactive map as you would with ipyleaflet.Map
.
m = geemap.Map()
To display it in a Jupyter notebook, simply ask for the object representation:
m
To customize the map, you can specify various keyword arguments, such as center
([lat, lon]), zoom
, width
, and height
. The default width
is 100%
, which takes up the entire cell width of the Jupyter notebook. The height
argument accepts a number or a string. If a number is provided, it represents the height of the map in pixels. If a string is provided, the string must be in the format of a number followed by px
, e.g., 600px
.
m = geemap.Map(center=[40, -100], zoom=4, height=600)
m
To hide a control, set control_name
to False
, e.g., draw_ctrl=False
.
m = geemap.Map(data_ctrl=False, toolbar_ctrl=False, draw_ctrl=False)
m
Adding basemaps¶
There are several ways to add basemaps to a map. You can specify the basemap to use in the basemap
keyword argument when creating the map. Alternatively, you can add basemap layers to the map using the add_basemap
method. Geemap has hundreds of built-in basemaps available that can be easily added to the map with only one line of code.
Create a map by specifying the basemap to use as follows. For example, the Esri.WorldImagery
basemap represents the Esri world imagery basemap.
m = geemap.Map(basemap="Esri.WorldImagery")
m
You can add as many basemaps as you like to the map. For example, the following code adds the OpenTopoMap
basemap to the map above:
m.add_basemap("Esri.WorldTopoMap")
m.add_basemap("OpenTopoMap")
Print out the first 10 basemaps:
basemaps = list(geemap.basemaps.keys())
len(geemap.basemaps)
basemaps[:10]
You can also change basemaps interactively using the basemap GUI.
Using Earth Engine data (30 mins)¶
Earth Engine data types (Image, ImageCollection, Geometry, Feature, FeatureCollection)¶
Earth Engine objects are server-side objects rather than client-side objects, which means that they are not stored locally on your computer. Similar to video streaming services (e.g., YouTube, Netflix, and Hulu), which store videos/movies on their servers, Earth Engine data are stored on the Earth Engine servers. We can stream geospatial data from Earth Engine on-the-fly without having to download the data just like we can watch videos from streaming services using a web browser without having to download the entire video to your computer.
- Image: the fundamental raster data type in Earth Engine.
- ImageCollection: a stack or time-series of images.
- Geometry: the fundamental vector data type in Earth Engine.
- Feature: a Geometry with attributes.
- FeatureCollection: a set of features.
Image¶
Raster data in Earth Engine are represented as Image objects. Images are composed of one or more bands and each band has its own name, data type, scale, mask and projection. Each image has metadata stored as a set of properties.
Loading Earth Engine images¶
image = ee.Image("USGS/SRTMGL1_003")
image
Visualizing Earth Engine images¶
m = geemap.Map(center=[21.79, 70.87], zoom=3)
image = ee.Image("USGS/SRTMGL1_003")
vis_params = {
"min": 0,
"max": 6000,
"palette": ["006633", "E5FFCC", "662A00", "D8D8D8", "F5F5F5"], # 'terrain'
}
m.add_layer(image, vis_params, "SRTM")
m
ImageCollection¶
An ImageCollection
is a stack or sequence of images. An ImageCollection
can be loaded by passing an Earth Engine asset ID into the ImageCollection
constructor. You can find ImageCollection
IDs in the Earth Engine Data Catalog.
Loading image collections¶
For example, to load the image collection of the Sentinel-2 surface reflectance:
collection = ee.ImageCollection("COPERNICUS/S2_SR")
Visualizing image collections¶
To visualize an Earth Engine ImageCollection, we need to convert an ImageCollection to an Image by compositing all the images in the collection to a single image representing, for example, the min, max, median, mean or standard deviation of the images. For example, to create a median value image from a collection, use the collection.median()
method. Let's create a median image from the Sentinel-2 surface reflectance collection:
m = geemap.Map()
collection = ee.ImageCollection("COPERNICUS/S2_SR")
image = collection.median()
vis = {
"min": 0.0,
"max": 3000,
"bands": ["B4", "B3", "B2"],
}
m.set_center(83.277, 17.7009, 12)
m.add_layer(image, vis, "Sentinel-2")
m
Filtering image collections¶
m = geemap.Map()
collection = (
ee.ImageCollection("COPERNICUS/S2_SR")
.filterDate("2021-01-01", "2022-01-01")
.filter(ee.Filter.lt("CLOUDY_PIXEL_PERCENTAGE", 5))
)
image = collection.median()
vis = {
"min": 0.0,
"max": 3000,
"bands": ["B4", "B3", "B2"],
}
m.set_center(83.277, 17.7009, 12)
m.add_layer(image, vis, "Sentinel-2")
m
FeatureCollection¶
A FeatureCollection is a collection of Features. A FeatureCollection is analogous to a GeoJSON FeatureCollection object, i.e., a collection of features with associated properties/attributes. Data contained in a shapefile can be represented as a FeatureCollection.
Loading feature collections¶
The Earth Engine Data Catalog hosts a variety of vector datasets (e.g,, US Census data, country boundaries, and more) as feature collections. You can find feature collection IDs by searching the data catalog. For example, to load the TIGER roads data by the U.S. Census Bureau:
m = geemap.Map()
fc = ee.FeatureCollection("TIGER/2016/Roads")
m.set_center(-73.9596, 40.7688, 12)
m.add_layer(fc, {}, "Census roads")
m
Filtering feature collections¶
m = geemap.Map()
states = ee.FeatureCollection("TIGER/2018/States")
fc = states.filter(ee.Filter.eq("NAME", "Louisiana"))
m.add_layer(fc, {}, "Louisiana")
m.center_object(fc, 7)
m
feat = fc.first()
feat.toDictionary()
m = geemap.Map()
states = ee.FeatureCollection("TIGER/2018/States")
fc = states.filter(ee.Filter.inList("NAME", ["California", "Oregon", "Washington"]))
m.add_layer(fc, {}, "West Coast")
m.center_object(fc, 5)
m
region = m.user_roi
if region is None:
region = ee.Geometry.BBox(-88.40, 29.88, -77.90, 35.39)
fc = ee.FeatureCollection("TIGER/2018/States").filterBounds(region)
m.add_layer(fc, {}, "Southeastern U.S.")
m.center_object(fc, 6)
Visualizing feature collections¶
m = geemap.Map(center=[40, -100], zoom=4)
states = ee.FeatureCollection("TIGER/2018/States")
m.add_layer(states, {}, "US States")
m
m = geemap.Map(center=[40, -100], zoom=4)
states = ee.FeatureCollection("TIGER/2018/States")
style = {"color": "0000ffff", "width": 2, "lineType": "solid", "fillColor": "FF000080"}
m.add_layer(states.style(**style), {}, "US States")
m
m = geemap.Map(center=[40, -100], zoom=4)
states = ee.FeatureCollection("TIGER/2018/States")
vis_params = {
"color": "000000",
"colorOpacity": 1,
"pointSize": 3,
"pointShape": "circle",
"width": 2,
"lineType": "solid",
"fillColorOpacity": 0.66,
}
palette = ["006633", "E5FFCC", "662A00", "D8D8D8", "F5F5F5"]
m.add_styled_vector(
states, column="NAME", palette=palette, layer_name="Styled vector", **vis_params
)
m
Earth Engine Data Catalog¶
The Earth Engine Data Catalog hosts a variety of geospatial datasets. As of October 2023, the catalog contains over 1,000 datasets with a total size of over 80 petabytes. Some notable datasets include: Landsat, Sentinel, MODIS, NAIP, etc. For a complete list of datasets in CSV or JSON formats, see the Earth Engine Datasets List.
Searching for datasets¶
The Earth Engine Data Catalog is searchable. You can search datasets by name, keyword, or tag. For example, enter "elevation" in the search box will filter the catalog to show only datasets containing "elevation" in their name, description, or tags. 52 datasets are returned for this search query. Scroll down the list to find the NASA SRTM Digital Elevation 30m dataset. On each dataset page, you can find the following information, including Dataset Availability, Dataset Provider, Earth Engine Snippet, Tags, Description, Code Example, and more (see {numref}ch03_gee_srtm
). One important piece of information is the Image/ImageCollection/FeatureCollection ID of each dataset, which is essential for accessing the dataset through the Earth Engine JavaScript or Python APIs.
m = geemap.Map()
m
dataset_xyz = ee.Image("USGS/SRTMGL1_003")
m.add_layer(dataset_xyz, {}, "USGS/SRTMGL1_003")
m = geemap.Map()
dem = ee.Image("USGS/SRTMGL1_003")
vis_params = {
"min": 0,
"max": 4000,
"palette": ["006633", "E5FFCC", "662A00", "D8D8D8", "F5F5F5"],
}
m.add_layer(dem, vis_params, "SRTM DEM")
m
Using the datasets module¶
from geemap.datasets import DATA
m = geemap.Map(center=[40, -100], zoom=4)
dataset = ee.Image(DATA.USGS_GAP_CONUS_2011)
m.add_layer(dataset, {}, "GAP CONUS")
m
from geemap.datasets import get_metadata
get_metadata(DATA.USGS_GAP_CONUS_2011)
Converting Earth Engine JavaScripts to Python¶
Find some Earth Engine JavaScript code that you want to convert to Python. For example, you can grab some sample code from the Earth Engine Documentation.
m = geemap.Map()
m
# Load an image.
image = ee.Image("LANDSAT/LC08/C02/T1_TOA/LC08_044034_20140318")
# Define the visualization parameters.
vizParams = {"bands": ["B5", "B4", "B3"], "min": 0, "max": 0.5, "gamma": [0.95, 1.1, 1]}
# Center the map and display the image.
m.set_center(-122.1899, 37.5010, 10)
# San Francisco Bay
m.add_layer(image, vizParams, "False color composite")
Exercise 1 - Creating cloud-free imagery¶
Create a cloud-free imagery of Texas for the year of 2022. You can use either Landsat 9 or Sentinel-2 imagery. Relevant Earth Engine assets:
m = geemap.Map(center=(40, -100), zoom=4)
dem = ee.Image("USGS/SRTMGL1_003")
landsat7 = ee.Image("LANDSAT/LE7_TOA_5YEAR/1999_2003").select(
["B1", "B2", "B3", "B4", "B5", "B7"]
)
states = ee.FeatureCollection("TIGER/2018/States")
vis_params = {
"min": 0,
"max": 4000,
"palette": ["006633", "E5FFCC", "662A00", "D8D8D8", "F5F5F5"],
}
m.add_layer(dem, vis_params, "SRTM DEM")
m.add_layer(
landsat7,
{"bands": ["B4", "B3", "B2"], "min": 20, "max": 200, "gamma": 2.0},
"Landsat 7",
)
m.add_layer(states, {}, "US States")
m.add("inspector")
m
Using the plotting tool¶
m = geemap.Map(center=[40, -100], zoom=4)
landsat7 = ee.Image("LANDSAT/LE7_TOA_5YEAR/1999_2003").select(
["B1", "B2", "B3", "B4", "B5", "B7"]
)
landsat_vis = {"bands": ["B4", "B3", "B2"], "gamma": 1.4}
m.add_layer(landsat7, landsat_vis, "Landsat")
hyperion = ee.ImageCollection("EO1/HYPERION").filter(
ee.Filter.date("2016-01-01", "2017-03-01")
)
hyperion_vis = {
"min": 1000.0,
"max": 14000.0,
"gamma": 2.5,
}
m.add_layer(hyperion, hyperion_vis, "Hyperion")
m.add_plot_gui()
m
Set plotting options for Landsat.
m.set_plot_options(add_marker_cluster=True, overlay=True)
Set plotting options for Hyperion.
m.set_plot_options(add_marker_cluster=True, plot_type="bar")
from geemap.legends import builtin_legends
for legend in builtin_legends:
print(legend)
Add NLCD WMS layer and legend to the map.
m = geemap.Map(center=[40, -100], zoom=4)
m.add_basemap("Esri.WorldImagery")
m.add_basemap("NLCD 2021 CONUS Land Cover")
m.add_legend(builtin_legend="NLCD", max_width="100px", height="455px")
m
Add NLCD Earth Engine layer and legend to the map.
m = geemap.Map(center=[40, -100], zoom=4)
m.add_basemap("Esri.WorldImagery")
nlcd = ee.Image("USGS/NLCD_RELEASES/2021_REL/NLCD/2021")
landcover = nlcd.select("landcover")
m.add_layer(landcover, {}, "NLCD Land Cover 2021")
m.add_legend(
title="NLCD Land Cover Classification", builtin_legend="NLCD", height="455px"
)
m
Custom legends¶
Add a custom legend by specifying the colors and labels.
m = geemap.Map(add_google_map=False)
keys = ["One", "Two", "Three", "Four", "etc"]
# colors can be defined using either hex code or RGB (0-255, 0-255, 0-255)
colors = ["#8DD3C7", "#FFFFB3", "#BEBADA", "#FB8072", "#80B1D3"]
# legend_colors = [(255, 0, 0), (127, 255, 0), (127, 18, 25), (36, 70, 180), (96, 68 123)]
m.add_legend(keys=keys, colors=colors, position="bottomright")
m
Add a custom legend by specifying a dictionary of colors and labels.
m = geemap.Map(center=[40, -100], zoom=4)
m.add_basemap("Esri.WorldImagery")
legend_dict = {
"11 Open Water": "466b9f",
"12 Perennial Ice/Snow": "d1def8",
"21 Developed, Open Space": "dec5c5",
"22 Developed, Low Intensity": "d99282",
"23 Developed, Medium Intensity": "eb0000",
"24 Developed High Intensity": "ab0000",
"31 Barren Land (Rock/Sand/Clay)": "b3ac9f",
"41 Deciduous Forest": "68ab5f",
"42 Evergreen Forest": "1c5f2c",
"43 Mixed Forest": "b5c58f",
"51 Dwarf Scrub": "af963c",
"52 Shrub/Scrub": "ccb879",
"71 Grassland/Herbaceous": "dfdfc2",
"72 Sedge/Herbaceous": "d1d182",
"73 Lichens": "a3cc51",
"74 Moss": "82ba9e",
"81 Pasture/Hay": "dcd939",
"82 Cultivated Crops": "ab6c28",
"90 Woody Wetlands": "b8d9eb",
"95 Emergent Herbaceous Wetlands": "6c9fb8",
}
nlcd = ee.Image("USGS/NLCD_RELEASES/2021_REL/NLCD/2021")
landcover = nlcd.select("landcover")
m.add_layer(landcover, {}, "NLCD Land Cover 2021")
m.add_legend(title="NLCD Land Cover Classification", legend_dict=legend_dict)
m
Creating color bars¶
Add a horizontal color bar.
m = geemap.Map()
dem = ee.Image("USGS/SRTMGL1_003")
vis_params = {
"min": 0,
"max": 4000,
"palette": ["006633", "E5FFCC", "662A00", "D8D8D8", "F5F5F5"],
}
m.add_layer(dem, vis_params, "SRTM DEM")
m.add_colorbar(vis_params, label="Elevation (m)", layer_name="SRTM DEM")
m
Add a vertical color bar.
m.add_colorbar(
vis_params,
label="Elevation (m)",
layer_name="SRTM DEM",
orientation="vertical",
max_width="100px",
)
Make the color bar background transparent.
m.add_colorbar(
vis_params,
label="Elevation (m)",
layer_name="SRTM DEM",
orientation="vertical",
max_width="100px",
transparent_bg=True,
)
m = geemap.Map()
m.split_map(left_layer="Esri.WorldTopoMap", right_layer="OpenTopoMap")
m
Create a split map with Earth Engine layers.
m = geemap.Map(center=(40, -100), zoom=4, height=600)
nlcd_2001 = ee.Image("USGS/NLCD_RELEASES/2019_REL/NLCD/2001").select("landcover")
nlcd_2019 = ee.Image("USGS/NLCD_RELEASES/2021_REL/NLCD/2021").select("landcover")
left_layer = geemap.ee_tile_layer(nlcd_2001, {}, "NLCD 2001")
right_layer = geemap.ee_tile_layer(nlcd_2019, {}, "NLCD 2021")
m.split_map(left_layer, right_layer)
m
Linked maps¶
Create a 2x2 linked map for visualizing Sentinel-2 imagery with different band combinations. Note that this feature does not work properly with Colab. Panning one map would not pan other maps.
image = (
ee.ImageCollection("COPERNICUS/S2")
.filterDate("2018-09-01", "2018-09-30")
.map(lambda img: img.divide(10000))
.median()
)
vis_params = [
{"bands": ["B4", "B3", "B2"], "min": 0, "max": 0.3, "gamma": 1.3},
{"bands": ["B8", "B11", "B4"], "min": 0, "max": 0.3, "gamma": 1.3},
{"bands": ["B8", "B4", "B3"], "min": 0, "max": 0.3, "gamma": 1.3},
{"bands": ["B12", "B12", "B4"], "min": 0, "max": 0.3, "gamma": 1.3},
]
labels = [
"Natural Color (B4/B3/B2)",
"Land/Water (B8/B11/B4)",
"Color Infrared (B8/B4/B3)",
"Vegetation (B12/B11/B4)",
]
geemap.linked_maps(
rows=2,
cols=2,
height="300px",
center=[38.4151, 21.2712],
zoom=12,
ee_objects=[image],
vis_params=vis_params,
labels=labels,
label_position="topright",
)
m = geemap.Map(center=[40, -100], zoom=4)
collection = ee.ImageCollection("USGS/NLCD_RELEASES/2019_REL/NLCD").select("landcover")
vis_params = {"bands": ["landcover"]}
years = collection.aggregate_array("system:index").getInfo()
years
Create a timeseries inspector for NLCD. Note that ipyleaflet has a bug with the SplitControl. You can't pan the map, which should be resolved in the next ipyleaflet release.
m.ts_inspector(
left_ts=collection,
right_ts=collection,
left_names=years,
right_names=years,
left_vis=vis_params,
right_vis=vis_params,
width="80px",
)
m
Time slider¶
Note that this feature may not work properly with Colab. Restart Colab runtime if the time slider does not work.
Create a map for visualizing MODIS vegetation data.
m = geemap.Map()
collection = (
ee.ImageCollection("MODIS/MCD43A4_006_NDVI")
.filter(ee.Filter.date("2018-06-01", "2018-07-01"))
.select("NDVI")
)
vis_params = {
"min": 0.0,
"max": 1.0,
"palette": "ndvi",
}
m.add_time_slider(collection, vis_params, time_interval=2)
m
Create a map for visualizing weather data.
m = geemap.Map()
collection = (
ee.ImageCollection("NOAA/GFS0P25")
.filterDate("2018-12-22", "2018-12-23")
.limit(24)
.select("temperature_2m_above_ground")
)
vis_params = {
"min": -40.0,
"max": 35.0,
"palette": ["blue", "purple", "cyan", "green", "yellow", "red"],
}
labels = [str(n).zfill(2) + ":00" for n in range(0, 24)]
m.add_time_slider(collection, vis_params, labels=labels, time_interval=1, opacity=0.8)
m
Visualizing Sentinel-2 imagery
m = geemap.Map(center=[37.75, -122.45], zoom=12)
collection = (
ee.ImageCollection("COPERNICUS/S2_SR")
.filterBounds(ee.Geometry.Point([-122.45, 37.75]))
.filterMetadata("CLOUDY_PIXEL_PERCENTAGE", "less_than", 10)
)
vis_params = {"min": 0, "max": 4000, "bands": ["B8", "B4", "B3"]}
m.add_time_slider(collection, vis_params)
m
Exercise 2 - Creating land cover maps with a legend¶
Create a split map for visualizing NLCD land cover change in Texas between 2001 and 2019. Add the NLCD legend to the map. Relevant Earth Engine assets:
import geemap
m = geemap.Map()
m
m = geemap.Map()
centroid = ee.Geometry.Point([-122.4439, 37.7538])
image = ee.ImageCollection("LANDSAT/LC08/C01/T1_SR").filterBounds(centroid).first()
vis = {"min": 0, "max": 3000, "bands": ["B5", "B4", "B3"]}
m.center_object(centroid, 8)
m.add_layer(image, vis, "Landsat-8")
m.add_layer(centroid, {}, "Centroid")
m
Check image properties.
image.propertyNames()
Check image property values.
image.toDictionary()
Get specific image properties.
image.get("CLOUD_COVER") # 0.05
Get image properties with easy-to-read time format.
props = geemap.image_props(image)
props
Compute image descriptive statistics.
stats = geemap.image_stats(image, scale=30)
stats
m = geemap.Map(center=[40, -100], zoom=4)
# Add NASA SRTM
dem = ee.Image("USGS/SRTMGL1_003")
dem_vis = {
"min": 0,
"max": 4000,
"palette": ["006633", "E5FFCC", "662A00", "D8D8D8", "F5F5F5"],
}
m.add_layer(dem, dem_vis, "SRTM DEM")
# Add 5-year Landsat TOA composite
landsat = ee.Image("LANDSAT/LE7_TOA_5YEAR/1999_2003")
landsat_vis = {"bands": ["B4", "B3", "B2"], "gamma": 1.4}
m.add_layer(landsat, landsat_vis, "Landsat", False)
# Add US Census States
states = ee.FeatureCollection("TIGER/2018/States")
style = {"fillColor": "00000000"}
m.add_layer(states.style(**style), {}, "US States")
m
Compute zonal statistics. In this case, we compute the mean elevation of each state and export the results to a CSV file.
out_dem_stats = "dem_stats.csv"
geemap.zonal_stats(
dem, states, out_dem_stats, stat_type="MEAN", scale=1000, return_fc=False
)
Display the csv file as a table.
geemap.csv_to_df(out_dem_stats).sort_values(by=["mean"], ascending=True)
Compute zonal statistics of mean spectral values of each state.
out_landsat_stats = "landsat_stats.csv"
geemap.zonal_stats(
landsat,
states,
out_landsat_stats,
stat_type="MEAN",
scale=1000,
return_fc=False,
)
geemap.csv_to_df(out_landsat_stats)
Zonal statistics by group¶
Compute zonal statistics by group. In this case, we compute the area of each land cover type in each state and export the results to a CSV file.
m = geemap.Map(center=[40, -100], zoom=4)
# Add NLCD data
dataset = ee.Image("USGS/NLCD_RELEASES/2021_REL/NLCD/2021")
landcover = dataset.select("landcover")
m.add_layer(landcover, {}, "NLCD 2021")
# Add US census states
states = ee.FeatureCollection("TIGER/2018/States")
style = {"fillColor": "00000000"}
m.add_layer(states.style(**style), {}, "US States")
# Add NLCD legend
m.add_legend(title="NLCD Land Cover", builtin_legend="NLCD")
m
Compute zonal statistics by group and convert the area unit from square meters to square kilometers.
nlcd_stats = "nlcd_stats.csv"
geemap.zonal_stats_by_group(
landcover,
states,
nlcd_stats,
stat_type="SUM",
denominator=1e6,
decimal_places=2,
)
geemap.csv_to_df(nlcd_stats)
Calculate the percentage of each land cover type in each state.
nlcd_stats = "nlcd_stats_pct.csv"
geemap.zonal_stats_by_group(
landcover,
states,
nlcd_stats,
stat_type="PERCENTAGE",
denominator=1e6,
decimal_places=2,
)
geemap.csv_to_df(nlcd_stats)
Zonal statistics with two images¶
The zonal statistics by zone algorithm is similar to the zonal statistics by group algorithm, but it takes an image as the zone input instead of a feature collection.
m = geemap.Map(center=[40, -100], zoom=4)
dem = ee.Image("USGS/3DEP/10m")
vis = {"min": 0, "max": 4000, "palette": "terrain"}
m.add_layer(dem, vis, "DEM")
m
landcover = ee.Image("USGS/NLCD_RELEASES/2021_REL/NLCD/2021").select("landcover")
m.add_layer(landcover, {}, "NLCD 2021")
m.add_legend(title="NLCD Land Cover Classification", builtin_legend="NLCD")
Computer the mean elevation of each land cover type.
stats = geemap.image_stats_by_zone(dem, landcover, reducer="MEAN")
stats
stats.to_csv("mean.csv", index=False)
Compute the standard deviation of each land cover type.
geemap.image_stats_by_zone(dem, landcover, out_csv="std.csv", reducer="STD")
geemap.csv_to_df("std.csv")
Exercise 3 - Zonal statistics¶
Find out which state has the highest mean temperature on in the United States on June 28, 2023. Relevant Earth Engine assets:
lat_grid = geemap.latitude_grid(step=5.0, west=-180, east=180, south=-85, north=85)
m = geemap.Map()
style = {"fillColor": "00000000"}
m.add_layer(lat_grid.style(**style), {}, "Latitude Grid")
m
df = geemap.ee_to_df(lat_grid)
df.head()
Create a longitudinal grid with a 5-degree interval.
lon_grid = geemap.longitude_grid(step=5.0, west=-180, east=180, south=-85, north=85)
m = geemap.Map()
style = {"fillColor": "00000000"}
m.add_layer(lon_grid.style(**style), {}, "Longitude Grid")
m
Create a rectangular grid with a 10-degree interval.
grid = geemap.latlon_grid(
lat_step=10, lon_step=10, west=-180, east=180, south=-85, north=85
)
m = geemap.Map()
style = {"fillColor": "00000000"}
m.add_layer(grid.style(**style), {}, "Coordinate Grid")
m
Creating fishnets¶
Create a fishnet based on an Earth Engine geometry.
m = geemap.Map()
m
Use the drawing tools to draw a polygon on the map above. If no polygon is drawn, the default polygon will be used.
roi = m.user_roi
if roi is None:
roi = ee.Geometry.BBox(-112.8089, 33.7306, -88.5951, 46.6244)
m.add_layer(roi, {}, "ROI")
m.center_object(roi)
Create a fishnet based on a user-drawn polygon with a 2-degree interval.
fishnet = geemap.fishnet(roi, h_interval=2.0, v_interval=2.0, delta=1)
style = {"color": "blue", "fillColor": "00000000"}
m.add_layer(fishnet.style(**style), {}, "Fishnet")
Create a new map.
m = geemap.Map()
m
Draw a polygon on the map.
roi = m.user_roi
if roi is None:
roi = ee.Geometry.Polygon(
[
[
[-64.602356, -1.127399],
[-68.821106, -12.625598],
[-60.647278, -22.498601],
[-47.815247, -21.111406],
[-43.860168, -8.913564],
[-54.582825, -0.775886],
[-60.823059, 0.454555],
[-64.602356, -1.127399],
]
]
)
m.add_layer(roi, {}, "ROI")
m.center_object(roi)
m
Create a fishnet based on a user-drawn polygon with specified number of rows and columns.
fishnet = geemap.fishnet(roi, rows=6, cols=8, delta=1)
style = {"color": "blue", "fillColor": "00000000"}
m.add_layer(fishnet.style(**style), {}, "Fishnet")
Land use and land cover change analysis¶
Forest cover mapping¶
We will use the Hansen Global Forest Change v1.10 (2000-2022) dataset.
dataset = ee.Image("UMD/hansen/global_forest_change_2022_v1_10")
dataset.bandNames()
Select the imagery for 2000.
m = geemap.Map()
first_bands = ["first_b50", "first_b40", "first_b30"]
first_image = dataset.select(first_bands)
m.add_layer(first_image, {"bands": first_bands, "gamma": 1.5}, "Landsat 2000")
Select the imagery for 2022.
last_bands = ["last_b50", "last_b40", "last_b30"]
last_image = dataset.select(last_bands)
m.add_layer(last_image, {"bands": last_bands, "gamma": 1.5}, "Landsat 2022")
Select the tree cover imagery for 2000.
treecover = dataset.select(["treecover2000"])
treeCoverVisParam = {"min": 0, "max": 100, "palette": ["black", "green"]}
name = "Tree cover (%)"
m.add_layer(treecover, treeCoverVisParam, name)
m.add_colorbar(treeCoverVisParam, label=name, layer_name=name)
m.add("layer_manager")
m
Extract tree cover 2000 by using the threshold of 10%.
threshold = 10
treecover_bin = treecover.gte(threshold).selfMask()
treeVisParam = {"palette": ["green"]}
m.add_layer(treecover_bin, treeVisParam, "Tree cover bin")
Forest loss and gain mapping¶
Visualize forest loss.
m = geemap.Map()
m.add_basemap("Esri.WorldImagery")
treeloss_year = dataset.select(["lossyear"])
treeLossVisParam = {"min": 0, "max": 22, "palette": ["yellow", "red"]}
layer_name = "Tree loss year"
m.add_layer(treeloss_year, treeLossVisParam, layer_name)
m.add_colorbar(treeLossVisParam, label=layer_name, layer_name=layer_name)
m.add("layer_manager")
m
Compare forest loss and gain.
m = geemap.Map()
m.add_basemap("Esri.WorldImagery")
treeloss = dataset.select(["loss"]).selfMask()
treegain = dataset.select(["gain"]).selfMask()
m.add_layer(treeloss, {"palette": "red"}, "Tree loss")
m.add_layer(treegain, {"palette": "yellow"}, "Tree gain")
m.add("layer_manager")
m
Zonal statistics by country¶
Compute zonal statistics to find out which country has the largest forest area in 2000.
Add a country boundary layer to the map.