Urban & Soil

BSI: Isolating Bare Soil From Everything Growing On It

Concept article · Updated · by Dr. Anant Kumar Pathak

Bare ground has a spectral fingerprint all its own — high in red and SWIR, low in NIR and blue — and BSI combines all four bands to separate exposed soil from vegetation and water in a single number.

How it reads the surface

The numerator plays soil's bright bands (SWIR1 + Red) against vegetation's (NIR + Blue), so exposed ground rises while canopy and water fall. Drawing on four bands makes BSI more robust than a two-band ratio at the tricky soil/vegetation edge.

The math behind it

BSI = ((SWIR1 + Red) − (NIR + Blue)) / ((SWIR1 + Red) + (NIR + Blue))

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

How to interpret the output

Typical range: −1 to +1

Higher = more exposed bare soil; low or negative = vegetated or water-covered ground.

Advertisement

The bands it needs

Sentinel-2SWIR1 B11 · Red B4 · NIR B8 · Blue B2
Landsat 8/9SWIR1 B6 · Red B4 · NIR B5 · Blue B2

Real-world use

BSI supports desertification and land-degradation monitoring, fallow-field detection in agriculture, and erosion and mining-footprint mapping.

Watch out for

Sensitive to soil type and moisture, and can confuse dry bare soil with some built-up materials.

Soil moisture and soil type shift BSI values, and some dry built-up materials mimic bare soil — for change work, compare BSI on similar-season, similar-moisture dates rather than across wet and dry periods.

Compute BSI on your own study area

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

Run this analysis in GISforus →

Frequently asked

What does a high BSI value mean?

Strongly exposed bare soil; vegetated and water-covered ground return low or negative values.

Does soil moisture affect BSI?

Yes — wet soil reads differently from dry, so compare dates with similar moisture and season for reliable change analysis.

Primary reference: Rikimaru, A., Roy, P.S. & Miyatake, S. (2002). Tropical forest cover density mapping. Tropical Ecology 43(1), 39–47.