Our latest additions to the Earth Engine Data Catalog

A summary of the new additions to the Earth Engine Data Catalog, and a glance at how fast it’s growing.

Google Earth
Google Earth and Earth Engine
5 min readNov 6, 2019

--

By Simon Ilyushchenko and Sulien O’Neill, on behalf of the Earth Engine Data team

The Earth Engine Data Catalog is an online repository of more than 600 datasets. This equates to 29 petabytes of data (that’s nearly 3X the size of all the photos uploaded to Google Photos in its first year of existence) with a growth rate of approximately one petabyte per month. All of the datasets in the catalog are updated automatically on a daily basis, and sometimes every 6 hours.

The Earth Engine Ingestion team maintains this data catalog to empower the Earth Engine community to devote less time to downloading data and managing infrastructure and more time on data analysis — pioneering new approaches to earth observation and science mapping, measuring, and monitoring our changing planet.

A snapshot of new datasets we’ve added since June of 2018.

  • Sentinel 5P: Daily concentrations of eight air pollutants. The “offline” collection can be used for historical trend analysis, while the “near real-time” collection relates current conditions.
  • Sentinel 2 SR: Land surface reflectance imagery for mapping landscapes. Being corrected for atmospheric effects, this imagery is well suited for change detection and time series analysis.
  • JAXA yearly SAR mosaics: Global annual synthetic aperture radar (SAR) mosaics. This data provides information about the structure and composition of Earth’s surface features.
  • Copernicus CORINE: Land cover maps for Europe representing 44 classes at 100m resolution. Data exist for 1990, 2000, 2006, and 2012.
  • Intertidal change: Global maps of tidal flats at 30m pixel resolution for 11 time-periods. These data are useful for studying changes in tidal flat extent.
  • Arctic and Antarctic DEMs: High quality Arctic and Antarctic surface elevation derived from optical stereo imagery. Datasets include full-extent mosaics and time series strips at 2m and 8m resolution.
  • USGS GAP/Landfire ecosystems: Detailed vegetation and land cover classification for the Conterminous U.S., Alaska, Hawaii, and Puerto Rico. Seamless mosaics represent vegetation information at the plant community level for the year 2011.
  • OpenLandMap data: A suite of data layers that define and describe global biomes, including monthly MODIS-based temperature and precipitation composites, soil characteristics, FAPAR, and potential vegetation type.
  • GOES-16 and GOES-17: Fire/hot spot characterization dataset providing near real-time monitoring of western hemisphere fire activity. Data layers include: fire temperature, fire area, fire radiative power, and a quality indicator.

A closer look at some of the new datasets

Sentinel 5-P TROPOMI

The Sentinel 5-P TROPOMI collection, one of the latest instruments in the joint European Commission and European Space Agency (ESA) Copernicus program, provides the ability to perform air quality analysis and visualizations on a wide range of pollutants like never before. You have access to analysis-ready columns of ozone, sulfur dioxide, nitrogen dioxide, carbon monoxide, formaldehyde, methane, tropospheric columns of ozone, vertical profiles of ozone, as well as cloud and aerosol information.

The daily global TROPOMI observations can be used for improving air quality forecasts as well as for monitoring concentrations of atmospheric constituents. The image below is a time series animation of nitrogen dioxide concentrations from July 1 to October 1, 2019 (during the time of the Amazon fires).

A time-series animation of nitrogen dioxide concentrations from July 1 to October 1, 2019 (during the time of the Amazon fires).

Sentinel-2 Surface Reflectance dataset (Level-2A)

New to the Sentinel-2 collection in Earth Engine is the Sentinel-2 Surface Reflectance dataset (Level-2A). This product provides Bottom Of Atmosphere (BOA) reflectance images, derived by the European Space Agency (ESA), from the associated Level-1C products. The wide-swath, high-resolution, multi-spectral imaging mission collects data that can be used for monitoring vegetation, soil and water cover, as well as observations of inland waterways and coastal areas. The Sentinel-2 mission has a revisit frequency of every 5 days at the equator.

The surface reflectance data product offers a higher-level data product, with the effects of the Earth’s atmosphere removed. Sentinel-2 Top of Atmosphere (left) and Sentinel-2 Surface Reflectance (right).

Additional outputs include Aerosol Optical Thickness (AOT) band, a Water Vapour (WV) band, and a Scene Classification Map (SCM), as well as Quality Indicators for cloud and snow probabilities at 60 m resolution.

GOES 16 (East) and Goes 17 (West) FDCF

One of the recent additions to our catalog is the GOES 16 (East) and Goes 17 (West) FDCF (Fire Data Classification Full disk). The GOES Fire product,(provided by NOAA & NASA from the GOES-R program) contains four images: one in the form of a fire mask and the other three with pixel values identifying fire temperature, fire area, and fire radiative power. The Earth Engine catalog has the GOES data from June 2019 to present.

CORINE Land Cover map

The Coordination of Information on the Environment (CORINE) Land Cover map (provided by the Copernicus Land Monitoring Service) is intended to provide consistent localized geographical information on the land cover across the European Union. The number of participating countries currently includes 33 (EEA) member countries and six cooperating countries (EEA39) with a total area of over 5.8 Mkm2.

The reference year of the first CLC inventory was 1990 and the first update created in 2000. Later, the update cycle has decreased to 6 years. Satellite imagery provides the geometrical and thematic basis for mapping with in-situ data as essential ancillary information. The basic technical parameters of CLC (i.e. 44 classes in nomenclature, 25 hectares minimum mapping unit (MMU), and 100 meters minimum mapping width) have not changed since the beginning; therefore, the results of the different inventories are comparable.

To get started, head over to the Earth Engine Data Catalog. If you’re new to Earth Engine, you’ll want to sign up first. Happy coding!

--

--