Unclassified Quantitative Symbology for Rasters (”Stretched” Colors)

Included in this tutorial

  • Stretched symbology defaults

  • Symbology options

  • Statistics options

  • Examining the histogram and changing the stretch type

  • Changing the color ramp

Software version in examples: ArcGIS Pro 2.9.1

Tutorial Data: This tutorial uses a 1/3 arc-second digital elevation model (DEM) available from the USGS to demonstrate. You can find and download DEM data for within the United States within the USGS National Map here.

Credits: L. Meisterlin (2022)

 

This tutorial demonstrates the options for applying a color ramp to quantitative values in a raster dataset, by “stretching” the ramp across the values’ range.

Related Tutorials:

 

Stretched, by default

For many raster datasets (depending on values, number of bands, and metadata), adding the raster to your map will result in a stretched symbology by default. This is true for a non-integer, single-band dataset used in the demonstration. In almost all GIS software, this default will use a black-to-white color ramp: darker shades applied to cells with lower values and lighter shades applied to cells with higher values.

the default stretched symbology of a raster

 

The Symbology Options

As with vector features, we access a raster’s symbology options through the Symbology panel. You can summon the Symbology panel by right-clicking on the layer’s name in the Contents panel and choosing Symbology.

accessing the Symbology panel by right-clicking on a layer’s name in the Contents panel

It is worth a moment to examine the default settings and options for raster symbols. Aside from the top two options (Primary Symbology: Stretch and the Band option, which is greyed-out here because the dataset is a single band), we will discuss them in the opposite order that they are presented in the interface.

the primary options for a stretched symbology

 

the Statistics options

The options under the Statistics tab allow you to choose which (and how) values are used in when stretching the color ramp across the range of values.

  • The default option is to stretch the color ramp from the minimum value of the full raster dataset to its maximum value, using the statistics (mean, standard deviation, etc) based on the distribution of these values.

  • Another common option is Dynamic Range Adjustment (DRA), which uses only the values in the currently visible area of the map. (DRA is further demonstrated in the Dynamic Range Adjustment for symbolizing rasters tutorial.)

  • Depending on your raster and analysis, you may also choose to assign the values used to create the stretched ramp and apply colors to each cell based on its value.

The demonstration below toggles between these options. Notice the change in the raster’s symbology as the dataset’s values are considered versus only the visible cells.

raster symbology statistics options

 

Examining the Histogram and Changing Stretch Type

The statistics options above are useful in combination with changing the stretch type to answer specific questions. Both sets of options, of course, benefit from examining the dataset’s histogram.

  • The default “Percent Clip” option removes the outer values from consideration when determining how to stretch the color ramp across the cell values. Here, the default setting is 0.5% from the minimum and the maximum. This is useful when your dataset has outliers that may skew representation. Examining the histogram helps to identify these potential outliers or the effect of clipping different percentages.

  • Within the histogram’s panel, you can also change the stretch type (not all options are present in the histogram panel). The demonstration below shows various options within this panel and within the Primary Symbology panel.

changing stretch types and examining the raster’s histogram

 

Changing the Color Ramp

Lastly, the demonstration below walks through changing some of the color ramp options, including customizing a color ramp based on the default option.

changing the color ramp options

 
Previous
Previous

Categorical (Unique Values) Symbology for Raster Datasets

Next
Next

Classified Quantitative Symbology for Raster Datasets