Setting up multiple imagery sources in CEO

Being able to view multiple sources of imagery is helpful for data collectors when they are analyzing landscapes. Different imagery sources can contribute to being able to accurately answer survey questions.

Institutional administrators should set up diverse imagery sources in order to support data collectors. Which specific imagery sources should be added to a Collect Earth Online (CEO) project depends on the survey questions being asked. Here we will discuss setting up a diversity of imagery tailored for your project.

🕵️‍♂️Consider your project needs

Think about the survey questions being asked in your CEO project. Are your questions asking about small scale phenomena or large scale patterns? If you are asking about small scale phenomena you will need high resolution images (1m-10m pixel size) for your data collectors to accurately answer the survey questions. Conversely if you are asking about broad scale patterns, more coarse resolution images will work (10m-30m pixel size).

Next, think about what time periods need to be available for data collection. Are you asking about multiple, specific time periods? If so, having multiple imagery sources with your specific dates pre-specified makes it easier for data collectors to answer survey questions. It also will help eliminate errors, for example if data collectors accidentally choose the wrong time period.

You should also think about what types of visualizations will be useful for your data collection. If you are looking at vegetation patterns, false color imagery using near infrared, red, and green light bands may be useful. If you are looking at agricultural fields, you may want to use false color imagery with shortwave infrared, near infrared, and green light bands specifically. Other band combinations or Synthetic Aperture Radar (SAR) are useful for examining other land cover and land use patterns.

Finally, consider if you need any specialty imagery. For example, this could include street names, vegetation height, or specialty calculations you have created.

🍱Adding multiple imagery sources

CEO’s built in imagery is a great place to start. By default, projects are able to use MapBox imagery and Planet imagery made available by Norway’s International Climate & Forests Initiative (NICFI). Mapbox provides cloud-free basemaps by compiling images from a variety of sources and dates, while Planet NICFI images are time-stamped, allowing users to access basemaps from a particular moment in time. Planet NICFI imagery is only available in the tropics and will by default show the most recent month available.

In addition to these built-in imagery sources, users can easily add imagery sources to their institution. After adding imagery to the institution, it becomes available for use in the institution’s projects. Note that institutional imagery is only visible to data collectors who are members of the institution.

One excellent option is Sentinel 1 and Sentinel 2 data, which comes from long-running public satellites (2014 and 2015 respectively). These imagery sources allow you to make data from last week or many years ago available to data collectors. The data is available at 10m, and therefore useful for many types of questions. Sentinel 2 data allows for multiple band combinations, while Sentinel 1 data is SAR data. Below are some pointers on adding Sentinel data to your institution. More detailed instructions are available in our manuals.

1. Select Sentinel 1 or Sentinel 2 from the institutional imagery addition menu.
2. Add a Title, which will be the displayed name of the imagery in any project the imagery is added to.
3. Add the Default Year and Default Month, which are the default year and month that will be displayed when the map loads. For the month, use integer format (1-12).
4. Select your Band Combination. For Sentinel 1, single and dual polarization options are available. For Sentinel 2, True Color, False Color Infrared, False Color Urban, Agriculture, Healthy Vegetation, and Short Wave Infrared are available. Data collectors will be able to switch the band combination in the data collection interface.
5. Add your Min and Max values. These are mapped to 0 and 255 respectively for visualization. They should be single numbers. For Sentinel 1, min = 0 and max = 0.3 is frequently used, while for Sentinel 2, min = 0 and max = 3000 is commonly used.
6. For Sentinel 2, choose your Cloud Score, which is the allowable cloud cover for the image. Values can be between 0-100 in multiples of 10.

Another great option if your project needs high resolution images in the tropics and you want to specify specific months and years for your data collectors is to add your own Planet NICFI key. This allows you to create multiple Planet NICFI imagery feeds for your project, so data collectors can easily switch between them. You will need to create an account for the Planet NICFI program. Once you have your key, here are some pointers on creating a Planet NICFI data stream with a custom date:

1. Select Planet NICFI from the institutional imagery addition menu.
2. Type in a Title.
3. Add in your Planet NICFI API key to the Access Token field. This can be accessed through your My Account page on the Planet website.
4. Choose the Default Time period you want to display. You can also choose the Default Band combination; both visible (RGB) and infrared false color are available.

You can repeat this process with the same API key to create multiple imagery streams with different default time periods.

If you need more custom imagery, you can connect any raster data in GEE, including both Image Assets and Image Collection Assets to your CEO institution. The platform includes a large Data Catalog of publicly hosted data, and also allows users to upload their own raster data. This is a fantastic way to add diverse imagery sources to your project. Many options at different spatial and temporal resolutions are available, along with specialty imagery like land cover classifications. Please see our recent blog post for more information on adding GEE raster data to your institution!

Another great option for more custom imagery is to connect your own WMS to your CEO institution. This is a great option if you have existing orthophotos, land cover model outputs, or other datasets that you would like to add to CEO. Please see our recent blog post, available in English, Spanish, and Portuguese for more information about adding this imagery to your institution.

Other great imagery options that CEO supports include Planet Monthly and Daily (paid), SecureWatch (paid), Bing Maps (key required), XYZ Tiles, and OpenStreetMaps. Planet Monthly, Planet Daily, and SecureWatch are high resolution near-real time imagery options that require a paid subscription. Bing Maps is a worldwide free high resolution datasets that requires a free account to generate an API key, much like the Planet NICFI option discussed above. XYZ Tiles is a format for tiled web mapping used by services including OpenStreetMaps. Adding OpenStreetMaps to your CEO project allows data collectors to view street names and other important information.

🌱Using KML to connect with other programs

In addition to all of these fantastic imagery options, CEO can export a Keyhole Markup Language (KML) file for any plot. The KML file contains information about the shape of your plot. Once downloaded, you can then use the KML information to view the plot in another program.

To download the plot KML, click on the Download Plot KML button in the data collection interface. Your web browser will ask you where you want to save your KML. Choose the appropriate location, then open it in your chosen program.

Any program that can import a KML file can be used. This includes Google Earth Pro, QGIS, ArcGIS, and many more. This great functionality in CEO allows you to leverage any of these other programs to further analyze your plot and support your data collection.

CEO would like to thank its ongoing funders FAO, NASA–USAID SERVIR, and SilvaCarbon, a US government program. Thanks also to CEO’s technology partners: Norway’s International Climate & Forests Initiative for funding open high-resolution data availability; Planet for providing high-resolution imagery; and the Google Earth Engine team for creating a platform for Earth science data and analysis.

Collect Earth Online is working constantly to improve the user experience, and your feedback is invaluable. If you have ideas to share, please write to support@collect.earth.

Thank you!