Advancing Geospatial Analysis in Snowflake with a New Data Type: GEOMETRY

GEOMETRY, a data type designed for those using local spatial reference systems (SRS), is now generally available in Snowflake. In addition, Snowflake has joined the Open Geospatial Consortium (OGC), an organization uniting over 500 businesses, government agencies, research organizations, and universities with the goal of making location information findable, accessible, interoperable, and reusable. 

Before this, Snowflake offered the GEOGRAPHY data type, which remains an excellent choice for customers using the most commonly used SRS relying on latitude and longitude coordinates on the earth’s surface—also referred to as the WGS 84. GEOGRAPHY data also uses spherical geometry for all the computations. In contrast, GEOMETRY uses planar geometry and supports thousands of projections; this is especially useful in representing local areas while minimizing distortion errors.

We know that mapping a spherical object, such as the earth, onto a flat surface like a map involves trade-offs. For instance, it’s impossible to preserve shapes of areas or objects and their relative sizes for the entire globe within a single SRS, which essentially means that preserving one characteristic will inevitably lead to the distortion of the other. When geographical information systems experts analyze spatial phenomena, they choose the most suitable coordinate system for their specific region and use case. 

That’s why we’ve introduced the GEOMETRY data type—to enable the usage of local SRS. It has a data structure and about 60 new spatial functions to give customers more flexibility in their geospatial analyses. Users can specify the SRS of data assigning a spatial reference identifier (SRID) when creating a GEOMETRY object and access information about the SRID by using the ST_SRID function. To enable conversion between different spatial reference systems, we’ve also started a private preview of the ST_TRANSFORM function.

In the example below, we’ve visualized the shape of the United Kingdom using three different spatial reference systems. Each SRS has unique characteristics, leading to similarities or significant differences in the displayed shapes.

Using GEOMETRY with its multiple SRIDs can be particularly beneficial in many areas including land surveying, urban planning, or resource management, where precise measurements are essential. It also allows better compatibility among different data sources that use specific SRIDs.

Want to learn more about GEOMETRY and how it differs from GEOGRAPHY? Check out our blog post Getting Started with GEOMETRY Data.

As Snowflake expands its support of industry-standard data types and libraries, we’ve joined the Open Geospatial Consortium (OGC) as a technical member.

The OGC is an international organization focused on developing geospatial standards and promoting interoperability in the geographic information ecosystem. Snowflake’s presence in the OGC helps us connect with fellow vendors in the geospatial space and better integrate with their products. Moving forward, we plan to engage in OGC working groups and initiatives to contribute to and benefit from the knowledge and expertise of the larger OGC community.

Nadine Alameh, CEO of the OGC, said, “I’m excited to welcome Snowflake as our newest Technical member at the Open Geospatial Consortium for two reasons: (1) their direct involvement in OGC’s cloud-native geospatial standardization activities will undoubtedly accelerate the uptake of those standards as well as the impact of the sharing and processing of geospatial data in the cloud; and (2) their engagement with the geospatial community via OGC underscores the value of geospatial in today’s connected world, truly helping Snowflake’s customers and ecosystem build on the value of incorporating geospatial/location in their data science workflows.”

For the latest on GEOMETRY data support, check out the Snowflake documentation.