What Is a Buffer in GIS?

A buffer in GIS is a polygon representing the area within a specified distance of a geographic feature. A buffer around a point creates an area surrounding that location, a buffer around a line creates a corridor along it, and a positive buffer around a polygon expands its area outward.

For example, a 500-metre buffer around a school represents locations that are within 500 metres of the school according to the distance model used by the GIS operation. A 100-metre buffer around a river creates a corridor extending 100 metres from its geometry.

Buffers are among the most common vector operations because many geographic questions begin with some version of “what is near this feature?” But a buffer is specifically a model of geometric distance. It should not automatically be interpreted as travel time, accessibility, impact, risk, or real-world service coverage.

Buffers at a glance

• Input geometry: Point; Typical buffer result: Area surrounding the point; Example use: Area within 1 km of a facility

• Input geometry: Line; Typical buffer result: Corridor along the line; Example use: 100 m zone around a river

• Input geometry: Polygon; Typical buffer result: Expanded polygon; Example use: 500 m zone around a protected area

• Input geometry: Polygon with negative distance; Typical buffer result: Smaller polygon; Example use: Area remaining after a 50 m inward setback

A standard buffer returns polygonal geometry because its purpose is to represent an area within a distance.

QGIS describes its Buffer operation as computing a buffer area around the features of an input vector layer using a fixed or data-defined distance.

A buffer turns distance into geometry

Suppose you have a point representing a health facility and choose a distance of 2 kilometres.

Conceptually, the buffer asks:

Which locations are no more than 2 kilometres from this point?

The result is an area surrounding the point.

For a single point on a suitable planar coordinate system, that area appears approximately circular:

        _________
     .-'         '-.
   .'               '.
  /         •         \
 |       facility      |
  \                   /
   '.               .'
     '-._________.-'

Every location inside the resulting polygon is within the chosen geometric distance of the input point, subject to the way distance and coordinates are handled by the implementation.

PostGIS defines ST_Buffer in essentially these geometric terms: the output covers points whose distance from the input geometry is less than or equal to the specified buffer distance.

That definition also explains why the result is not merely a visual circle drawn around a map symbol. The buffer is new geometry that can itself be queried, intersected, measured, published, or used in later analysis.

Point buffers create areas around locations

Point buffers are probably the easiest form to recognise.

If each point represents a school, a 1-kilometre buffer can create a polygon around each school representing geometric proximity within that distance.

Potential questions include:

• Which buildings lie within 1 kilometre of a school?

• Which observations occur within 5 kilometres of a monitoring station?

• Which parcels fall within 500 metres of a facility?

• Which settlements are close to a proposed site under a straight-line distance definition?

The phrase under a straight-line distance definition matters.

A buffer does not know whether a road connects the locations, whether a river blocks travel, whether a gate is open, or whether someone is allowed to cross a particular boundary. It is constructing geometry from distance, not modelling movement through the real world.

Line buffers create corridors

Buffering a line creates an area along its length.

A 100-metre buffer around a river, for example, creates a corridor extending approximately 100 metres from the river geometry on both sides when a conventional two-sided buffer is used.

Similar workflows might involve:

• zones around roads;

• setbacks around pipelines;

• areas near coastlines;

• corridors around transmission lines;

• proximity zones around trails.

The shape at corners and line endings is not completely predetermined. Buffer implementations commonly expose options controlling end caps and joins.

A line ending can, for example, be represented with a rounded, flat, or square-style cap. Corners can use round, bevelled, or mitred joins.

Those parameters change the exact polygon produced, so the nominal distance is not the only setting that can matter geometrically.

For most introductory analysis, the defaults may be sufficient. For engineering, design, cadastral, or other precision-sensitive workflows, the chosen buffer style should be deliberate.

Polygon buffers expand an existing area

A positive buffer around a polygon extends its boundary outward.

Suppose a protected-area polygon represents the legally designated area itself, but an analysis requires locations within 2 kilometres of its boundary.

A positive buffer can create an expanded polygon covering the protected area and the surrounding distance zone.

This is structurally different from buffering a point because the starting feature already represents an area.

The buffer does not replace the original meaning of that polygon. It creates a derived area based on distance from it.

That distinction matters when the output is later published. A polygon produced by buffering a protected area is not itself necessarily part of the legally protected area; it is an analytical construction around it.

Negative buffers move polygon boundaries inward

Some GIS implementations allow a negative buffer distance for polygons.

Instead of expanding the geometry, the operation shrinks it inward.

For example:

Original polygon
+--------------------+
|                    |
|                    |
|                    |
+--------------------+

Negative buffer
    +------------+
    |            |
    |            |
    +------------+

This can represent a geometric setback from a polygon boundary or help construct an interior area that excludes the outer edge.

If the negative distance is large enough, small or narrow polygons can disappear completely because no area remains after the inward offset.

Negative buffering therefore deserves more care than simply thinking of it as “buffering backwards”. The later What Is a Negative Buffer? article can treat its geometry and failure cases in detail.

Multiple buffers can overlap

Suppose five facilities are located close together and each receives a 1-kilometre buffer.

The resulting polygons may overlap.

That creates an important decision:

Should those buffer areas remain separate, or should their overlaps be dissolved into a combined coverage area?

If each buffer needs to remain associated with its original facility, keeping separate polygons can be useful.

If the question is simply:

Which areas fall within 1 kilometre of at least one facility?

then dissolving the overlapping buffers may produce a more appropriate representation.

These are not merely different visual styles. They create different feature structures and support different questions.

The planned Dissolved vs Overlapping Buffers article owns that distinction in detail. The underlying grouping operation is covered separately in What Does Dissolve Do in GIS?.

Buffer distance depends on how distance is represented

A buffer needs two things:

1. input geometry;

2. a distance.

But the number alone is meaningless without units and a distance model.

If a tool interprets coordinates in metres, a distance value of 500 may mean 500 metres.

If it is operating directly in a coordinate system whose units are degrees, simply entering 500 obviously does not mean the same thing.

Different GIS libraries also handle geographic coordinates differently. Some require an appropriate projected CRS for meaningful metric buffers, while others provide geography-aware buffering that internally chooses a suitable planar representation or otherwise accounts for the curved Earth.

PostGIS, for example, distinguishes between geometry buffering—where distance follows the units of the spatial reference system—and geography buffering, for which it chooses a planar spatial reference internally before calculating the result. PostGIS documents those behaviours explicitly.

The practical lesson is not that every buffer must use the same projection. It is that the units and distance model must match the question.

The supporting article Why Buffer Distance Depends on the CRS should own that problem in depth.

A buffer is not an isochrone

A buffer can answer:

What is within 5 kilometres geometrically?

It cannot, by itself, answer:

What can I reach within 15 minutes by road?

Those are different spatial models.

A buffer expands outward according to distance from the input geometry. It does not know about a transport network.

An isochrone or network-based service area can account for paths through a network and, depending on the model, factors such as:

• roads;

• walking paths;

• permitted movement;

• travel speeds;

• turn restrictions;

• bridges;

• network barriers.

Consider two settlements that are each 3 kilometres from a hospital in straight-line distance. One may have a direct road connection while the other sits across a river with the nearest bridge many kilometres away.

Both may fall inside the same geometric buffer while having very different real accessibility.

That is why a buffer should not be labelled a “15-minute service area” merely because a distance was chosen as an approximate proxy for travel.

The dedicated Buffer vs Isochrone article owns that distinction.

A buffer does not prove influence, exposure, or risk

Distance is often useful precisely because geographic processes are spatially related. But the existence of a buffer does not establish what happens within it.

A 1-kilometre buffer around an industrial facility tells you which locations satisfy a geometric-distance criterion.

It does not, without additional evidence, establish that:

• pollutants reach every point inside the buffer;

• locations outside it have no exposure;

• everyone inside experiences the same risk;

• the facility's effects stop at the buffer boundary.

Likewise, a buffer around a river is not automatically a floodplain, and a buffer around a road is not automatically an air-pollution exposure zone.

Those interpretations require models, observations, regulations, or domain-specific assumptions beyond the buffer operation itself.

A good map should distinguish the geometric rule used to construct the buffer from the real-world phenomenon being investigated.

Variable-distance buffers use different distances for different features

A buffer distance does not always have to be identical across a layer.

Suppose each facility has an attribute called:

buffer_metres

One facility might use 500 metres, another 1,000 metres, and another 2,000 metres.

Some GIS tools allow the buffer distance to be calculated from such an attribute or expression. QGIS, for example, supports data-defined distances in its standard Buffer algorithm.

This can be useful when the distance itself has a meaningful basis in the data.

But a variable buffer is only as defensible as the values used to create it. If those distances are arbitrary or derived from an unsuitable proxy, varying them feature by feature does not make the analysis more accurate.

A standard line buffer usually extends to both sides

A conventional buffer around a line normally creates area on both sides.

Sometimes that is not what the question requires.

For example, a riverbank regulation might concern only one side of a mapped line, or a road-related operation might specifically require the left or right side according to the line's direction.

GIS software can provide single-sided buffers for such cases. QGIS, for example, exposes left- and right-side line buffering separately from its general Buffer operation.

This is an important specialised form, but it introduces questions about line direction, side selection, joins, and geometry semantics that do not need to be folded into the basic definition.

The dedicated Single-Sided vs Two-Sided Buffers article should own those details.

Buffering creates derived geometry

A buffer result did not exist in the source data.

If the original dataset contains hospitals, a 5-kilometre buffer layer contains areas constructed from those hospital locations and a chosen distance.

This means the result should preserve enough context to answer:

• What features were buffered?

• What distance was used?

• In what units?

• What spatial reference or distance model was used?

• Were buffers dissolved?

• Were distances constant or feature-specific?

• Were special cap, join, or side settings applied?

Those parameters affect what the resulting polygons mean.

A buffer layer saved without that context can look authoritative while leaving a later reader unable to reconstruct why its boundaries fall where they do.

When should you use a buffer?

A buffer is a strong fit when your spatial question genuinely uses geometric distance.

Examples include:

Which parcels fall within 500 metres of this proposed site?

What land lies within 100 metres of this river geometry?

Which observations occur within 10 kilometres of these monitoring stations?

What polygon represents a 50-metre geometric setback from this boundary?

In each case, the output you need is an area defined by distance from existing geometry.

A buffer is less suitable when the intended question depends on:

• travel time;

• route distance;

• network accessibility;

• administrative membership;

• physical process modelling;

• observed impact;

• statistical association.

Those may still use buffers as one analytical component, but the buffer should not be mistaken for the phenomenon itself.

The most important question is what the distance means

Buffering is mechanically simple: take geometry, choose a distance, and construct an area around it.

The analytical decision is harder.

Before creating the buffer, ask:

Why this distance?

Is 500 metres defined by legislation? Is it an analytical threshold? Is it a convenient search radius? Does it represent a measurement assumption? Is it merely being used as a rough proxy for something that actually depends on a transport network or environmental process?

Then ask:

How should that distance be measured?

The answer determines whether the coordinate system, distance method, or specialised analysis needs more attention.

A buffer is most useful when those questions have clear answers. It then becomes exactly what it should be: a transparent geometric representation of locations within a defined distance of a feature.

References

1. QGIS Documentation — Buffer. Documents QGIS's standard vector Buffer operation, including fixed and data-defined distances, polygon output, negative polygon buffers, segment settings, cap and join styles, and dissolve-related options.

2. PostGIS — ST_Buffer. Defines a buffer as polygonal geometry covering locations within a specified distance of an input geometry or geography and documents differences in distance handling, negative buffers, and buffer styling.