Vegetation

EVI: The Vegetation Index Built for Dense, Hazy Canopies

Concept article · Updated · by Dr. Anant Kumar Pathak

When NDVI flat-lines over a rainforest, EVI keeps climbing. The Enhanced Vegetation Index was designed by the MODIS science team specifically to stay responsive where NDVI saturates, and to shrug off the atmospheric haze and soil glare that quietly distort it.

How it works

EVI keeps NDVI's red–near-infrared contrast but bolts on two corrections. A soil-adjustment term in the denominator dampens the influence of a bright or dark background, and a blue-band term subtracts the aerosol scattering that inflates red reflectance on hazy days. The coefficients (6, 7.5 and the gain of 2.5) come from radiative-transfer modelling, not guesswork.

Because the denominator no longer collapses toward a fixed ceiling, EVI keeps separating "80% canopy" from "95% canopy" — exactly the range where NDVI stops telling you anything.

The formula

EVI = 2.5 × (NIR − Red) / (NIR + 6 × Red − 7.5 × Blue + 1)

where NIR is near-infrared reflectance, Red is red reflectance, Blue is blue reflectance.

Reading the values

Typical range: ≈ −1 to +1 (typically 0–1 for vegetation)

Higher = greener, denser canopy. Stays responsive where NDVI plateaus above ~0.6, so it is preferred in tropical forest and high-biomass work.

Advertisement

Which bands to use

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

In practice

In tropical and high-biomass regions — oil-palm concessions, Amazon monitoring, peak-season maize — EVI is the default over NDVI. It is also the backbone of the global MODIS vegetation-product record, so using it keeps your work comparable with decades of published phenology studies.

Where it struggles

Relies on the blue band — the noisiest and most aerosol-sensitive channel — so it demands good atmospheric correction to be reliable.

EVI's Achilles heel is the blue band: it is the noisiest channel on any sensor and the most sensitive to residual atmosphere. If your imagery is not well atmospherically corrected, EVI can be less reliable than a plain NDVI — correct first, then choose EVI.

Compute EVI on your own study area

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

Run this analysis in GISforus →

Frequently asked

EVI or NDVI — which should I use?

Use NDVI for quick, sparse-to-moderate vegetation and long historical comparisons; use EVI for dense canopy, tropical forest or hazy scenes where NDVI saturates or atmosphere is a concern.

Why does EVI need the blue band?

The blue band is used to estimate and subtract aerosol scattering, which otherwise inflates red reflectance and biases the index on hazy days.

Primary reference: Huete, A., Didan, K., Miura, T., Rodriguez, E.P., Gao, X. & Ferreira, L.G. (2002). Remote Sensing of Environment 83(1–2), 195–213.