Why Do Two Boundary Datasets Not Line Up?

When two boundary datasets do not line up, the discrepancy may be a coordinate problem, a positional-accuracy problem, a time/version difference or a genuine difference in what the boundary represents. Do not snap one dataset to the other until you know which of those explanations applies.

The spatial pattern of the mismatch is the best first clue.

If the whole layer is shifted, suspect the reference system

A nearly uniform offset across the entire dataset often points to a CRS or datum issue.

Check whether both layers have known coordinate reference systems and whether those definitions are correct rather than merely present. Although two layers may both display in the same map because the software reprojects them on the fly, one could still have been assigned the wrong source CRS.

If the shift is systematic and similar everywhere, investigate:

• wrong CRS assignment;

• wrong datum or reference frame;

• missing geographic transformation;

• incorrect axis order;

• coordinates interpreted in the wrong units.

Do not “fix” a uniform shift by editing vertices manually. If the reference system is wrong, manual edits destroy otherwise coherent geometry.

If the difference varies locally, compare source accuracy and scale

Suppose two coastlines overlap closely in some places and diverge by tens or hundreds of metres elsewhere; that pattern is less consistent with one simple coordinate offset.

The datasets may have been created from different source imagery, survey methods or map scales. One may be a high-resolution administrative dataset; the other may be a generalised national or global boundary intended for small-scale maps.

Generalisation removes vertices and smooths shapes so a boundary remains readable at smaller scales, which is not necessarily an error. Overlaying a 1:5,000 cadastral boundary with a 1:10,000,000 world dataset can therefore reveal large differences even when both are appropriate for their intended uses.

Inspect metadata for nominal scale, resolution, positional accuracy and source methodology where available.

If the mismatch follows particular administrative areas, check the date

Boundaries change.

Municipalities merge, districts split, disputed areas are represented differently, coastlines are updated and census geography is revised between enumeration cycles, so two datasets can be individually correct for different dates and still fail to align.

Look for version dates and stable geographic identifiers. If names match but codes or versions differ, do not assume the geometry should be identical.

This is especially important for joins. Attribute data from one census year may not be directly joinable to boundaries from another if administrative units changed in between.

Why Geographic Joins Fail When Boundary Versions Differ focuses on that problem.

If the boundaries disagree in meaning, neither dataset may be wrong

A “boundary” is not always one universal object.

Possible differences include:

• legal boundary versus cartographic representation;

• coastline at different tidal definitions;

• land boundary versus statistical reporting area;

• official national view versus internationally harmonised dataset;

• disputed territory included, excluded or separately represented;

• administrative boundary versus service-delivery boundary.

If the two datasets encode different concepts, forcing geometric coincidence would remove information rather than repair it.

The metadata should tell you what each dataset claims to represent. If it does not, provenance becomes part of the quality problem.

Look at where the boundaries disagree

A useful diagnostic is to calculate or visually inspect the symmetric difference between the two layers. The pattern can separate broad classes of cause:

Mismatch pattern · More likely explanation

• Nearly constant shift — CRS/datum/transform issue

• Small irregular deviations everywhere — Different source accuracy or generalisation

• Changes concentrated in certain districts — Version or administrative change

• Coastline-only differences — Source resolution or coastline definition

• Disputed areas differ substantially — Political/legal representation

• One dataset consistently simpler — Generalisation or intended map scale

This does not prove the cause, but it tells you what evidence to inspect next.

Do not use snapping as the first response

Snapping can make boundaries coincide, but it cannot tell you which boundary is authoritative or whether coincidence is appropriate. If one dataset is known to be authoritative for the intended workflow and the other is meant to conform to it, controlled snapping or conflation may be justified; in that case, record the source, tolerance and changed features.

If both datasets are independent representations, preserve them and document the discrepancy instead of silently forcing one onto the other.

Verify with control evidence

Before choosing a corrective action, compare:

1. CRS definitions and transformations;

2. dates and versions;

3. source scale or resolution;

4. authoritative status;

5. known control points or surveyed boundaries;

6. definitions of the geographic units.

Then test several locations rather than one convenient overlap.

Misalignment is a diagnosis, not yet a defect

Two boundary datasets that do not line up are telling you that they were created under different assumptions, sources or reference systems. Sometimes one is wrong, sometimes one is more accurate, and sometimes both are correct for different purposes.

The defensible workflow is to identify whether the difference is positional, temporal or definitional before editing any geometry.

Related content

• Why Geographic Joins Fail When Boundary Versions Differ — version mismatch and joins

• Valid Geometry vs Accurate Geometry — structural validity versus positional truth

• What Is Spatial Data Provenance? — tracing where boundary data came from