Precision

In a geodatabase, precision describes how your data will be stored on your system. The geodatabase stores coordinates as integers. ArcGIS uses the precision value to convert the numbers with decimals to integers—the precision value determines how many decimal places will be preserved when a coordinate is converted to an integer. Up to nine decimal places can be preserved.

When you're working with your data in ArcMap or viewing it in ArcCatalog, precision is used to convert the coordinates back to real numbers (numbers with decimals).

The precision value is used to convert back and forth between the numbers with decimals displayed in ArcMap and ArcCatalog and the integers stored in the geodatabase.

Together with the coordinate values, the precision value defines the spatial domain of your data. Increasing the precision value decreases the spatial extent and, conversely, decreasing the precision value increases it. As the precision value increases, so too does the data resolution that can be stored in the geodatabase.

For example, a precision of 1,000 will store data at a higher resolution than a precision of 100. If your data is stored in meters, a precision of 1,000 will support data stored at millimeter accuracy. A precision of 100 will support data stored at centimeter accuracy.

Does higher precision equal higher accuracy?

It is important to understand that the precision value is not a measure of the accuracy of your data. Accuracy is the measure of how closely your data matches the true values in the real world. Using a higher precision value will not increase the accuracy of your data. For example, if your data has 1-meter accuracy, assigning a precision value of 1,000 will not give your data millimeter-level accuracy.

When you create a standalone feature class or feature dataset, ArcCatalog will provide either a placeholder value or a default value for precision. In both cases, the provided precision value will probably be higher than you need for your data. You can accept the precision value or you can modify it if you need to fine-tune your database.

After a feature class or feature dataset has been created, you cannot change its precision.