Removing slivers after splitting a polygon with a line

Bye-Bye Slivers: Cleaning Up After Polygon Splits in GIS

Ever split a polygon in GIS and ended up with those annoying little slivers? Yeah, me too. Those tiny, skinny polygons are like the crumbs after a cookie – unwanted byproducts of the splitting process. They pop up because of slight data imperfections or just the limitations of the software we’re using. These sliver polygons don’t actually represent anything real, and honestly, they can mess up your spatial analysis and make your maps look… well, not great. So, let’s dive into how to find and ditch these pesky things after a polygon split, so your data stays clean and accurate.

What Are Sliver Polygons Anyway?

Think of a sliver polygon as that awkward little strip of land left over when two property lines almost, but don’t quite, meet. They’re small, they’re narrow, and they usually hang out along the edges of other polygons. You’ll often see them after doing things like Intersect or Union operations. Typically, their area is ridiculously small compared to how long their perimeter is – imagine a super stretched-out rubber band. Basically, they’re errors we need to get rid of. Back in 1978, some researcher named Michael Goodchild even found that slivers made up most of the polygons in some GIS projects! Crazy, right?

So, how do these things even happen?

• Human Error: When digitizing maps manually, it’s easy to make tiny mistakes, especially when tracing complex shapes.
• Data Differences: If your data comes from different sources or uses different coordinate systems, things might not line up perfectly.
• GIS Operations: Splitting, buffering, and overlaying polygons can sometimes expose or even create these slivers. It’s just a side effect of the math!
• Forgetting to Snap: If you don’t use snapping tools when you’re editing, you can end up with little gaps or overlaps. Trust me, I’ve been there.

Hunting Down the Culprits: Identifying Slivers

Alright, before we can get rid of them, we need to find them. Here’s how GIS software can help:

• Eyeball It: For smaller datasets, you can just visually scan the map for those skinny polygons along the borders. It’s tedious, but sometimes it works.
• Attribute Sleuthing: Select polygons based on their area, perimeter, or shape. Slivers usually have a tiny area but a surprisingly long perimeter. Think long and skinny!
• Thinness Ratio: There is a formula for this. It’s T = 4π(A/P²). A thinness ratio of 1 is a perfect circle and a thinness ratio of 0 is a line .
• Topology Time: Use topology rules to find gaps, overlaps, and those dreaded small polygons. Topology defines how points, lines, and polygons share the same space.

The Extermination Plan: Removing Slivers

Okay, time to get our hands dirty. There are several ways to get rid of slivers, depending on your GIS software and how bad the problem is:

Example Time: The “Eliminate” Tool in Action (ArcGIS)

The “Eliminate” tool in ArcGIS is a go-to for sliver removal. Here’s how it works:

Proactive Protection: Preventing Slivers in the First Place

Honestly, the best way to deal with slivers is to stop them from happening in the first place. Here’s my advice:

• Snap, Snap, Snap: I can’t say it enough: always use snapping tools!
• Tolerance is Key: Use appropriate fuzzy tolerances during overlay operations.
• Validate Your Data: Regularly check your data for topological errors.
• Go High-Res: Use the highest resolution data you can get your hands on.
• Keep it Planar: Enforce planar topology to make sure everything exists on a flat, two-dimensional plane.

The End of the Sliver Story

Sliver polygons are just a fact of life in GIS, especially when you’re splitting polygons. But by understanding why they happen, knowing how to find them, and using the right tools to remove them, you can keep your data clean and accurate. And remember, a little prevention goes a long way! Trust me, your future self will thank you.