Earth Engine | Google Cloud
Google Earth Engine Analyze satellite imagery and geospatial data at planetary scale Improve sustainability and climate resilience decision-making with Earth Engine's curated geospatial data catalog and the large-scale computing and advanced AI of Google Cloud. Go to console Contact sales Product highlights A catalog of 90+ petabytes of analysis-ready geospatial data Analyze Earth Observation data and run Machine Learning on it Work in the JavaScript Code Editor or in Python Explore deeper connectivity across Earth Engine and Cloud Features Earth Engine in BigQuery Earth Engine capabilities are now available in BigQuery, making advanced geospatial analysis of satellite imagery accessible to the SQL community.
The ST_RegionStats() function, now in Preview, allows you to run a new BigQuery geography function to efficiently extract statistics from raster data within specified geographic boundaries Earth Engine datasets in Analytics Hub give you access to a growing collection of Earth Engine datasets directly within BigQuery, simplifying data discovery and access Learn how to use Earth Engine in BigQuery Earth Engine and BigQuery interoperability Earth Engine focuses on image (raster) processing, whereas BigQuery is optimized for processing large tabular datasets.
Using BigQuery with Earth Engine, users get the best of both worlds. Earth Engine users can now easily move data between Earth Engine and BigQuery: Use the ee.FeatureCollection.loadBigQueryTable() function to load data directly from a BigQuery table without transforming it within BigQuery Use the ee.FeatureCollection.runBigQuery() function to trigger BigQuery computation to evaluate a SQL query directly against BigQuery tables and retrieve the results as an Earth Engine FeatureCollection Use the `Export.table.toBigQuery()`function to write the results of Earth Engine computation to BigQuery Learn how to use the Earth Engine to BigQuery connector 90+ petabytes of analysis-ready geospatial data The Earth Engine catalog is one of the largest publicly available data catalogs, with 90+ petabytes of analysis-ready satellite imagery and 1,000+ curated geospatial datasets.
It includes 50+ years of historical imagery, updated and expanded daily, at resolutions as fine as one meter per pixel. Examples include Landsat , MODIS and Sentinel , the National Agriculture Imagery Program (NAIP) , climate and weather data, geophysical data, including terrain, land cover, and cropland data. The catalog offers data on the entire planet, allowing users to understand earth changes relevant to their sustainability goals.
Explore Earth Engine datasets Powerful computation platform at scale Google Cloud empowers everyone to run large-scale parallel processing using many thousands of computers. Combining Earth Engine’s data catalog with Google Cloud’s computational capability and data analytics tools, makes Earth Engine a revolutionary platform to analyze and visualize earth data at scale. Faster access, processing, and analysis of data means faster innovations, informed decisions, and viable solutions.
The U.S. Forest Service, for example, manages 193 million acres of U.S. forest lands, reducing the time to complete mission critical tasks from months to hours with Earth Engine’s superior data catalog and computational scalability. 1:36 Code Editor The Earth Engine Code Editor is a web-based coding environment designed to make developing complex geospatial workflows fast and easy with the following elements: JavaScript Code Editor Map display for visualizing geospatial datasets API reference documentation (Docs tab) Git-based Script Manager (Scripts tab) Console output (Console tab) Task Manager (Tasks tab) to handle long-running queries Interactive map query (Inspector tab) Search of the data archive or saved scripts Geometry drawing tools Learn more about Earth Engine Code Editor Earth Engine Python API and Colab experience (geemap) Earth Engine Python API allows users to make use of Python tools for machine learning and analysis, including those for geospatial workloads, like Cloud Optimized GeoTiffs and GeoPandas.
The geemap Python library is supported in Earth Engine, for visual workflows in Python, like panning, zooming, and drawing map polygons for zonal statistics. Xarray, a well-known Python package, enables working with multidimensional arrays. Xee , its Earth Engine integration, lets users work with Earth Engine ImageCollections as Xarray datasets. Learn about recent Earth Engine Python API developments Earth Engine Machine Learning Earth Engine has built-in capabilities to allow users to train and use ML models for common scenarios with easy-to-use APIs.
For instance, you can use a random forest algorithm to classify land in an area of interest. If you'd prefer to use a Deep Neural Network, you can also train a TensorFlow or PyTorch model, deploy it to Gemini Enterprise Agent Platform and get predictions from within the Earth Engine Code Editor. Learn how to get predictions from hosted models on Gemini Enterprise Agent Platform Import your own data for analysis with the Earth Engine data catalog Users can import their own data (images and tables) and combine them with datasets from the Earth Engine data catalog to derive insights.
Using the Asset Manager in Code Editor or command line interface (CLI) , georeferenced raster datasets in GeoTIFF or TFRecord format and tabular data in Shapefile or CSV format can be imported to build data products, create models, and develop unique solutions to accelerate sustainability efforts. Learn how to upload and manage geospatial datasets in Code Editor Export your results for integration into other systems If you're training a TensorFlow model or want to run hydrology simulations outside Earth Engine you might want to get data out of Earth Engine into another system .
The Earth Engine export API does the heavy lifting and our data extraction methods help solve scaling issues and work with frameworks like Apache Beam, Spark, or Dask. Our Python client library comes bundled with client-side logic to convert between Earth Engine objects and NumPy, Pandas, and GeoPandas types. Learn how to export images, map tiles, tables and video from Earth Engine Earth Engine Apps For code-free, interactive visualizations, Earth Engine Apps are dynamic, shareable user interfaces for Earth Engine analyses.
With Earth Engine Apps , developers can use simple UI elements to leverage Earth Engine's data catalog and analytical power, allowing stakeholders to interact with their data and delivering insights into the hands of decision-makers. Learn more about Engine Apps Cloud Score+ Cloud Score+ solves the issue of cloud cover in Sentinel-2 satellite data. It is a comprehensive QA score, powered by deep learning, which provides a per-pixel “usability” score to mask or weight observations based on overall quality.
Learn more about Cloud Score+ Dynamic World Dynamic World is a global, near-real-time Land Cover dataset at a 10m resolution, powered by machine learning. It gives unprecedented detail about land use and helps make accurate predictions and effective sustainability plans. Learn more about Dynamic World How It Works Reference analysis-ready datasets in Earth Engine's petabyte-scale data catalog. Run raster and vector data analysis in parallel on thousands of machines.
Iteratively test and tweak results in Code Editor, by writing custom Python code with Earth Engine's client library or by extracting results from a model hosted in Gemini Enterprise Agent Platform. Share results in BigQuery or in Earth Engine Apps. View documentation Solve for common geospatial use cases at scale Common Uses Sustainable sourcing Enable global supply chain transparency and traceability Sustainable supply chains are business critical.
Earth Engine helps businesses analyze land cover and use at sourcing sites to highlight deforestation risk in their supply chains. The EC JRC global map of forest cover
The rest of this page is on the live site.