This tutorial has been updated and published on the Google Earth Outreach website:
Please go there for the latest version. (March 2013)
Google Earth Engine is a planetary-scale platform for environmental data analysis. It brings together over 40 years of historical and current global satellite imagery, and provides the tools and computational power necessary to analyze and mine that vast data warehouse. Current applications include: detecting deforestation, classifying land cover and land cover change, estimating forest biomass and carbon, and mapping the world’s roadless areas.
This tutorial will introduce you to Google Earth Engine and its basic functionality, including exploring the Data Catalog and viewing datasets in the Workspace.
Let’s explore the Home page and find out a bit of what Earth Engine can do.
At the top of the page is a search bar, where you can search for places or datasets. For example, entering “Landsat” brings up datasets with Landsat in their name, description, or tags, while entering “Brazil” brings up locations with Brazil in their name. In the top right, there is a Sign in link, where Earth Engine partners can sign in.
The introductory text on the Home page gives an overview of Earth Engine as “a planetary-scale platform for environmental data and analysis.” It also provides links to product videos, news items, and other resources.
Below the introductory text is the Featured gallery, where you can quickly find examples of some of the best and latest analysis products produced by Google Earth Engine and the organizations using it. These include links to view massive datasets using the Google Earth plugin (available for Windows and Mac).
Note: If you don’t have the Google Earth browser plugin installed, you can get it here.
Other links in the Featured gallery will display timelapses of environmental change on massive scales. For most of the timelapses, clicking the link will show you an introductory video, like the one for the Drying of the Aral Sea shown below.
To explore the timelapse map, click the “Explore Map” button indicated above and you’ll see the zoomable, timelapse map shown below. On the timelapse maps, you can zoom in, pan the map, pause the timelapse, and select the playback speed, as indicated below.
The Data Catalog lists the datasets available for viewing and analysis in Google Earth Engine.
On the Data Catalog page you will see a list of Popular Tags, linking to datasets that have those tags applied. Below that is a list of various datatypes and multi-day mosaics, including brief descriptions of, and direct links to, a handful of the available datasets. These lists show or link to most of the datasets and mosaics available in Google Earth Engine. To access ALL available datasets, use the search bar at the top of the page.
This page shows details about the selected dataset, including its name, a brief description, a sample image, and information such as which dates are available (Landsat 5 stopped sending data at the end of 2011; use Landsat 7 or MODIS data for more recent imagery), the provider’s name, and any tags for that dataset. There is also a blue Open in workspace button which can be used to add the dataset to your current workspace (more on that below).
The Workspace is where you manage, analyze, and visualize datasets in Google Earth Engine.
On the Workspace page, you will see a map on the right, and space for a list of data layers on the left. Unless you have already added a dataset to your workspace, your Data list will be empty, and the map will show the Google Maps terrain layer, as shown above.
As a reminder for navigating the Google Maps interface, here are some basics. You can move (pan) around the map by clicking and dragging anywhere on the map. To zoom in and out there are several techniques. You can always use the [+] and [-] zoom buttons and the zoom slider on the map to zoom in and out. You can also double-click anywhere on the map to zoom in. If your pointing device (mouse or track pad) has a right button, you can double-right-click on the map to zoom out. If you have a touch-screen device, you may be able to zoom with a pinch gesture, and if you have a mouse with a scroll wheel, the easiest way to zoom is to simply turn the wheel. To change the map background use the buttons in the upper right of the map to select either Map view or Satellite view. When selecting Map view, a checkbox will appear below the Map button to turn on/off Terrain instead of the usual road-map view. When selecting Satellite, a checkbox will appear below the Satellite button allowing you to turn on/off the Labels (borders, countries, cities, water bodies, etc.).
Now let’s view some data in the Workspace...
You will see that the dataset is listed (MCD43A4...) in the Data layer list in the left-hand panel, and that the data is visible on the map.
Note that some datasets can only be shown at certain zoom levels, and not at others. For example, if you are zoomed all the way out to a global view and try to see a Landsat 7 dataset, it will not be visible on the map. Don’t worry, it’s not broken! A yellow bar appears at the top of the map saying that you need to zoom in to view the data.
Now we’ll adjust some settings to explore the data in more detail.
The Layer Settings allow you to customize a variety of parameters, including the date(s) for which data is shown.
Note: For “Classified Raster” type data layers, the settings required are different. See the Viewing Classified Rasters section below.
You can view multiple data layers on your map at once by adding additional datasets.
There are several ways to add additional data layers in your Workspace. The first method is to return to the Data catalog, select another dataset, and use its Open in Workspace button. This will add the dataset to your workspace, as a layer above your current data layer(s). Note that the new layer will show on top of the previous layers on your map. See below for changing the order of the layers.
Another way to add additional datasets is directly from the search bar your Workspace. To start searching for a dataset to add, do one of the options below:
All three of these options will allow you to type your query in the Search bar, and select a dataset to add as a layer.
When you add a new layer to your Workspace, “Raster” type datasets will come in as a simple layer, but “Classified Raster” type datasets require a bit of set up before you can view them (see the Viewing Classified Rasters section below). The screenshot below shows both types of datasets as results for the search “m”.
You can also add the same dataset twice, as two separate layers in your Workspace. One reason to do this would be to view two different time slices of the same dataset, to view change over time. For more on this, see the “Visualizing change over time” section below.
When you have more than one dataset visible on your map, the one listed at the top of the Data layers list will be drawn on top of those below it. To change the ordering, use your mouse to click on the drag point to the left of the dataset name in the list, and drag it up or down in the list.
If you wish to remove a data layer from your Workspace...
Now that you know the basics, let’s explore a few of the more powerful things you can do in the Google Earth Engine workspace. In the sections below we’ll show you how to setting up and view classified rasters, adjust a layer’s visualization parameters, and visualize change over time.
Classified Raster type data layers require a little more setup in order to view them. You will need to select the year to show, and set up classes with names and colors to represent each class. For example, the MCD12Q1 classified rasters represent 5 different systems for classifying land cover type. Each of these datasets is annual (ranging from 2001 to 2009), and divides the Earth into different land cover classes. More information about each of these classification systems may be found on the USGS Distributed Archive site.
Let’s set up a Classified Raster data layer...
When you add the classified raster, a Classes section appeared in the left-hand panel. You can use this to add classes and assign them colors and names, or you can do it in the Layer Settings dialog for the classified raster layer. These two techniques are described in the instructions and image below.
Classes that have not been assigned colors will not appear on the image. Classes may be removed from the image by clicking the X that appears next to a Class when you move your mouse over the class name.
Once you add classes to your workspace for each of the classes in the raster dataset, it will look something like this:
In the Layer Settings dialog for most data layers, you will see a “Visualization Parameters” link. Click on it to reveal a number of advanced visualization settings. Each dataset has different default values, which are shown when you first click the link, but you can modify them to change how you visualize the dataset.
The first row of parameters are Min, Max, Gain, Bias, and Gamma. These parameters let you modify how data values are visualized. You may either set Min and Max, or you may set Gain and/or Bias.
Min represents the value to represent as decimal value 0 and Max represents the value to represent as decimal value 255. The values below Min will also be drawn with value 0, and the values above Max will also be drawn with value 255. Values between Min and Max will be scaled linearly, so that the middle of the range will be assigned value 122.
For example, the Shuttle Radar Topography Mission (SRTM) dataset contains values that represent elevation in meters, from -425 m to 8806 m. To visualize the dataset with a good detail in most parts of the world, you might want to represent 0 meters as black and represent 3000 meters and above as white, so set the Min to 0 and Max to 3000. To pick out mountains, or better see variation in high elevation areas above 3000 meters, you can set the Min to 3000 and the Max to 8806.
The image below shows the SRTM dataset with Min = 0 and Max = 3000, showing Puget Sound and highlighting Mount Rainier (4,392 m tall) as the white spot.
An alternative way to alter how the values in a dataset map onto visualization values is the change the gain and bias. Each value is multiplied by the gain, and increased by the bias. For example, the SRTM values, which range between -415.0 and 8806 can be compressed to between 0 and 255 by multiplying by 0.02765 (set the gain to 0.02765). adding 11.47 (set the bias to 11.47).
Gamma represents the relationship between a value and the luminance used to represent it. Roughly speaking, increasing gamma increases the intensity of values in the middle of the visualization range.
When you see an image on the web, you are generally seeing a combination of red, green, and blue pixels (RGB). In Earth Engine, these are separated into “bands”: the red band contains the red values for each pixel, the blue band contains the blue values for each pixel, and the green band contains the green values for each pixel. These bands are then combined to form the image you see on the screen.
Many Earth Engine datasets include more than three bands. For example, Landsat 7 images have 8 bands. Three bands roughly match red, green, and blue, and others represent infrared light, or thermal energy. Each band has a name. In the case of Landsat, the blue band is named 10, the green band is named 20, and the red band is named 30. To see an image that looks like how we typically see the aerial imagery, Earth Engine maps bands 30, 20, 10 onto R,G, B, respectively.
However, mapping different bands onto R, G, and B can create some interesting and useful effects. For example, mapping bands 40, 30, and 20 onto R, G, and B creates a “false color” image in which vegetation is highlighted and displayed in red.
The Bands input field provides a place where you can tell Earth Engine which bands of a dataset you would like to represent as red, green, and blue. To do this, list the band names in the RGB order, separated by commas. For example, to see a false color image, type 40, 30, 20 into the Bands (R, G, B) input field, as shown below:
A palette allows you to assign colors to the range of values in a dataset. A palette is a series of comma separated hexidecimal color values. Providing two values sets the color of the lowest value and highest value of the dataset. For example the SRTM digital elevation model is displayed in shades of gray by default. To display it in shades of red instead, where the lowest elevation points are black, and the highest elevation points are dark red, enter 000000,FF0000 into the palette box (it looks better if you set Min to 0 and Max to 3000). FF0000 is the hexidecimal value that is high (FF) on red and low (00) on green and blue. The 000000 value is low on red, green, and blue. To make the low elevations white instead, use the palette FFFFFF,FF0000.
Adding an additional color values to the palette will divide the color range into two areas: beginning to midpoint, and midpoint to end. The colors in these ranges will be scaled from the beginning of each range to its end. Adding more colors will increase the number of color ranges. Try visualizing SRTM with the palette FFFFFF,00FF00,FF0000. To see a complicated palette, open an NDVI dataset (type NDVI into the search field) and open its Visualization Parameters. The image below shows NDVI around Sacramento, California.
One of the interesting things you can do in Google Earth Engine is to visualize changes over time. To do this, you will need to add the same dataset to your Workspace as two separate layers, but set them to show different time slices. The example below will show you how to visualize the rapid urban expansion of Las Vegas, Nevada.
There are a number of things to look out for and be aware of as you explore the data in Google Earth Engine, some related to the way Earth Engine works, and some implicit in the datasets. Below are some of the more common things you may run into.
Google Earth Engine has more advanced features such as classifying land cover, downloading datasets, and the ability to build your own data analysis algorithms. If you are interested in these features you will need to sign in, which is currently limited to organizations in our Trusted Tester program. To apply to be a trusted tester, please send a brief description of your organization and how you would like to use Google Earth Engine to: email@example.com. Once you are a trusted tester, you will be able to access additional tutorials.
Feedback for us?
If you have any feedback on the functionality or user interface, please let us know, so that we can be sure to take it into account as we continue to develop and improve Google Earth Engine. The best way to provide us feedback is through the Send Feedback link in the upper right of every Earth Engine page.