For those in the agriculture, environmental, or other land-use industries, the ability to view and analyze several strata of surface data is necessary to make critical decisions for your business. With Studio's Raster Tiling system, you have access to a powerful, fast, and easy-to-use raster processing workflow.

The Santa Monica Mountains following the Woolsey Fire.

What is Raster Data?

Raster data, commonly derived from remote sensors such as satellites, are collections of "images" containing multiple wavelengths of light, with high bit depth data stored in each pixel.

Studio is equipped with the Raster Layer, a tool to explore massive, petabyte-scale image collections from your web browser.

To learn how our Raster Tile layer works under the hood, be sure to check out the blog post we released at Raster Tile's launch.

Raster Tiles Overview

Raster Tiles can be added to Studio from the Data Catalog, or imported as a Cloud-Optimized GeoTIFF (COG) by referencing standardized Spatio-Temporal Asset Catalog (STAC) metadata.

Raster Tiles in the Data Catalog

If you are new to Raster Tiles (or are an experienced user who wants to assess Studio's raster capabilities) the Data Catalog is the best place to start. While the catalog is diverse and growing, the "Remote Sensing" section contains several invaluable resources at no cost to you.

Currently, the Data Catalog contains four remote sensing datasets, each with its value and purpose:

Import Custom Raster Data

Raster layers can reference user-provided, custom Cloud-Optimized GeoTIFFs (COG) by providing standardized Spatio-Temporal Asset Catalog (STAC) metadata.

1. Open Add Data to Map dialog, navigate the Tilesets tab, then select Raster Tile.
2. Enter a URL to a STAC metadata file into the Tileset Metadata field then click Add Data.

Color Rescaling

You may find that your Raster Layer provides poor visibility while viewing certain areas of the Earth's surface. Due to factors such as image quality, the type of terrain being studied, and the time and season at which the image was captured, there is no perfect, one-size-fits-all color scale configuration. Instead, users are expected to iterate on the color scale to apply an optimal contrast setting.

Studio makes this process painless — simply drag the color scale sliders and instantly see results in real-time. For finer control, enter exact numbers for color scale parameters.

Rendering Options (Spectral Bands)

Satellites like Landsat-8 and Sentinel-2 have specialized sensors that capture around a dozen different wavelengths of light reflected off the earth. These wavelengths can reveal data critical for scientific analyses.

For example, many remote sensing datasets also contain a near-infrared band, which is valuable for vegetation analyses because near-infrared light reflects off the chlorophyll in plants. Areas with denser vegetation have more near-infrared light that is captured by the satellite’s image sensor.

Filtering

With GPU-based pixel filtering, you can hide any pixels whose index value falls outside the desired range.

Split Map Modes

Analyzing surface data often involves drawing conclusions through comparison. Considering the availability of both infrared and near-infrared bands, users are expected to compare two images from the same raster tileset. This is easily achieved by duplicating the layer, then using a split map mode.

Use Case Examples

Find examples of common use cases represented in Studio below:

• Agriculture
• Emergency Management

Raster Tiles for Agriculture

Raster data plays a critical role in agriculture, helping feed millions through data-optimized decisions. Farming and water organizations rely on remote data to understand everything from the health of crops as they mature throughout the season, the optimal fertilizer blend for optimal biomass yield, the appropriate irrigation and rotation strategy, the impacts of droughts, and much more.

Agriculture: Analyzing Uneven Growth

Disproportionate emergence, uneven heights, and other abnormalities can be caused by a variety of factors such as the delta in soil temperature, seeding depth, residue distribution, soil crusting, etc.

Researchers measured the response of corn when 25, 50, or 75 percent of the plants were planted at either 10 or 21 days following the first plant date. Surprising results revealed a grain yield drop by 6 to 7 percent with a delayed planting of 10 days. Learn more at the LANDSAT missions page.

MSAVI = (2 _ NIR + 1 – sqrt((2 _ NIR + 1)² – 8 * (NIR - R))) / 2

As crops mature, it is suggested to switch to Normalized Difference Vegetation Index (NDVI).

NDVI = (NIR - R)/(NIR + R)

Emergency Management

In our rapidly changing environment, remote sensing data is more important than ever for preventing and evaluating natural disasters. To show an example of emergency management we will hone in on the Woolsey Fire, a wildfire that burned Los Angeles and Ventura Counties in California. The Woolsey Fire reached 100% containment on November 21st, 2018.

Embedded Example

Explore the embedded example below. This map was created entirely with free data available on the Data Catalog.

More Information

The above use case examples were completely constructed in Studio using data freely available via our growing Data Catalog.

• Raster Layer
• Add tiled datasets
• Data catalog
• Hex Tiles visualization
• Split Map Mode

If you are not sure if your intended use case is supported, take a second to join our community Slack channel where we frequently engage in public conversations.