Reclassifying Rasters
Included in this tutorial:
Accessing the Reclassify tool
Reviewing the tool’s options and parameters
Three examples and reviewing their results: Reclassifying by Quantile, Reclassifying with Manual ranges, and Creating a Binary Raster with reclassification
Software version in examples: ArcGIS Pro 3.0.0
Tutorial Data: The tutorial includes demonstration with sample data available here.
Credits: L. Meisterlin and Varisa Tanti (2022)
Reclassifying raster datasets is the process of producing a new raster dataset based on an existing raster layer, with values corresponding to a classification method that you establish. While spatial resolution is maintained, the granularity of the attribute information (cell values) is diminished in the process in favor of grouping attribute values into desired classes.
This tutorial demonstrates three reclassification with three classification methods—quantile, manual entry, and binary. The example input raster in all three examples was produced by converting the Tract layer in the Tutorial Data Standard Data Package to a raster layer, using the values its the Shape_Length field (as an arbitrary numeric variable).
Related: This tool requires either the Spatial Analyst Extension or the 3D Analyst extension in ArcGIS Pro. To verify your software’s extensions, see Which Esri Extensions Is My Software Licensed to Use?
Accessing the Reclassify Tool
You can access the Reclassify tool by clicking the Geoprocessing toolbox in the main ribbon, and navigating through Spatial Analysis Tools > Reclass > Reclassify. You can also search for “Reclassify” in the search bar.
the Reclassify tool in the geoprocessing pane
Reclassify Tool Options & Parameters
In the Reclassify Tool dialogue box, you will see these options: input raster, reclassifying field, reclassification, and output cell size.
the inputs of the Reclassify tool
Here’s a brief description of the different inputs:
Input raster: The input raster to be reclassified
Reclassifying field: Field denoting the values that will be reclassified. This is relevant for multi-band rasters, which include values from more than one “band” within a composite raster dataset. In a single-band raster, the only option will be the value of the cells in the raster dataset.
Reclassification: This table will display ranges of values (”Start” and “End”) to be reclassified with the “New” value, after these values are established by the user. The table is explored further below, with examples.
Classify (button): To use different classification methods and a different number of classes, click the “Classify” button below the Reclassification table. The options of the Classify menu should be fairly familiar. They are the same as those presented when using a classified symbology for raster datasets.
Unique (button): To reclassify unique values within the input raster (rather than classified values), click the “Unique” button below the Reclassification table. The options in the resulting menu will allow you to establish a one-to-one reclassification.
Output raster: The location and name of the output raster.
Reclassification: 3 Examples
Example 1: Reclassifying by Quantile
The Classify Menu
To reclassify by quantile, click the “Classify” button below the Reclassification table.
In the Classify menu, select “Quantile” from the Method drop-down (demonstrated in the screenshot on the left below).
Select the number of classes in the “Classes” drop-down menu (demonstrated in the screenshot on the right below). In the example, we are reclassifying the values in the input raster by quartile (four classes).
As with other instances where the Classify menu appears, you can also examine the histogram, and compare your class break values to the overall distribution of values.
Select ‘Quantile’ from the Method drop-down menu.
Select the number of classes in the Classes drop-down menu.
Click “OK” to close the Classify menu and add your classification values to the Reclassification table. The table should now reflect the numeric values of your classification ranges (under “Start” and “End” per class).
Assigning “New” Reclassified Values for the Output Raster
In the “New” column, the software will be populated with default values. The values in this column will be assigned to cells in the output raster corresponding to the input raster’s cell value.
To reverse the order of the new values, click “Reverse New Values” (above the table, on the right).
To manually assign new values, you can type within the cells in the New column by double-clicking each value. Similarly, the “NODATA” value can also be changed (for example, to zero).
As you make changes in this table, note that you can save your reclassification table (click the purple 3.5-disk icon) as well as load an existing table (click the yellow folder icon).
The Reclassification table reflecting the values of the quartile classification of the input raster value and the corresponding values in the new, output raster.
Running the Reclassify tool & Examining Results
When you are ready to process the tool, click “Run.”
The results of running our example quartile reclassification are shown in the screenshot below. Notice that the new raster dataset (called “Reclass_Feat4”) has only four values (1-4).
The resulting raster dataset after reclassifying the input by quartile
Example 2: Reclassifying by Manual Breaks
To reclassify by manual breaks, return to the Classify menu (click the “Classify” button below the Reclassification table).
With a manual classification, specify the number of classes first (in the Classes drop down), then select “Manual Interval” under the Method drop-down menu. (This is demonstrated in the screenshot on the left, below.) If you change the method to “Manual Interval” first, you will not be able to change the number of classes.
In our example, we are using five classes.
Specify the Class Break values
Manually specify the upper limit of each class, by double-clicking the values listed under “Upper value” and typing. (This is demonstrated in the screenshot on the right, below.)
The Method drop-down menu, under Classify. In this example, we are choosing “Manual Interval.”
Double-clicking the values in the “Upper value” column to modify the reclassification ranges manually
Click “OK” to close the Classify menu and add your classification values to the Reclassification table. The table should now reflect the numeric values of your classification ranges (under “Start” and “End” per class).
Assigning New (Reclassified) Values, Running the tool, & Examining Results
As with the Quantile reclassification example, next assign the new (reclassified) values to be assigned based on your manual ranges.
When you are satisfied with your choices, click “Run” to execute the tool and produce the output raster.
In the example below, we see the reclassified raster has five new values (ranging from zero to four).
the resulting raster dataset after reclassifying the input into five classes manually
Example 3: Creating a Binary Raster with Reclassification
You can use the Reclassify tool to create a binary raster, with values of 1 or 0.
Option 1: Using a manual classification
In this example, we will reclassify all input cells with a value as 1 in the output raster and all cells without a value (as NODATA in the input) as 0 in the output raster.
To do this, we can use a manual classification (following the steps above) with one class. Then, we can specify the “Upper value” as the maximum value within the input raster. This is demonstrated in the screenshot on the left, below.
As a result, our Reclassification table will reflect one class ranging from the minimum value to the maximum value. We can assign a new value of 1 to this entire range. This is demonstrated in the screenshot on the top right, below.
We can then assign a new value of 0 to all cells without values (NODATA), which is demonstrated in the screenshot on the bottom right, below.
changing the number of classes to “1” and the Classification Method to “Manual Interval.”
the Reclassification table with two rows: the top row contains the full range of input cell values; the bottom row includes cells of no value.
double-clicking in the “New” column to add “0” as the new value, and clicking the save icon
When you are ready to process the tool, click “Run” and examine your results.
resulting binary raster, generated by assigning all cells with values in the input raster a value of 1 and all other cells a value of 0
Option 2: Reclassifying unique values
You can also isolate a single value in your input raster and assign it a value of 1 (or any other new value) in the reclassified output. By giving all other cells a value of zero or “NoData”, we can use this approach to create binary rasters from specific values.
To specify unique values for reclassification, click the “Unique” button. With the unique values listed, you can manually enter new values as described above. (See the screenshots below.)
the “Unique” button below the Reclassification table allows you to manually enter the new cell values.
Double-click on the values in the New column to enter the new cell values. Here, we want to isolate the cells with values of “1.”
After executing the Reclassify tool, the results of our example include a new binary raster layer with cell values of 1 where the input also includes cell values of 1. All other cells (values 2 through 5 in the input layer) have a value of zero in the new raster.
the resulting binary raster layer, generated by reclassifying (keeping) values of 1 in the input and reclassifying all other values to zero