Urban & Soil

NDBI: Mapping the Built Environment From Space

Concept article · Updated · by Dr. Anant Kumar Pathak

Concrete and asphalt do something vegetation does not — they reflect more shortwave-infrared than near-infrared. NDBI turns that reversal into a first-order map of where the built environment is.

What it actually measures

Built-up and impervious surfaces push NDBI positive; vegetation and water go negative. It is the mirror image of NDVI's band logic, which is why the two are so often differenced together to isolate urban land from everything else.

The formula

NDBI = (SWIR1 − NIR) / (SWIR1 + NIR)

where SWIR1 is shortwave-infrared 1 reflectance (~1.6 µm), NIR is near-infrared reflectance.

Reading the values

Typical range: −1 to +1

> 0 built-up / impervious · < 0 vegetation and water. Bare soil sits close to built-up values.

Advertisement

Which bands to use

Sentinel-2SWIR1 B11 · NIR B8
Landsat 8/9SWIR1 B6 · NIR B5

In practice

Use it for rapid urban-extent mapping, impervious-surface estimation, and monitoring city expansion across dates.

Where it struggles

Bare soil and built-up behave similarly in SWIR/NIR, so NDBI alone over-maps soil as urban — combine it with NDVI/MNDWI (see IBI).

NDBI's blind spot is bare soil, which behaves almost identically in SWIR/NIR and gets over-mapped as built-up. In any scene with fallow fields or drylands, combine it with NDVI and a water mask — or use IBI, which does exactly that.

Compute NDBI on your own study area

Skip the code. Draw or upload a boundary and Spatial Research Suite runs NDBI on live Sentinel-2 or Landsat imagery — with cloud masking, exports and citations built in.

Run this analysis in GISforus →

Frequently asked

Why does NDBI confuse bare soil with buildings?

Dry bare soil has a SWIR/NIR response close to impervious surfaces, so both score positive; combining NDBI with NDVI/MNDWI (as IBI does) resolves it.

What NDBI value means built-up?

Positive values typically indicate built-up or impervious surfaces, but validate the threshold locally.

Primary reference: Zha, Y., Gao, J. & Ni, S. (2003). Use of normalized difference built-up index in automatically mapping urban areas from TM imagery. International Journal of Remote Sensing 24(3), 583–594.