MSAVI2: Soil Correction Without Guessing the L Factor
SAVI works — if you pick the right soil-adjustment constant. MSAVI2's trick is that it derives that adjustment from the image itself, so you never have to guess a value that is wrong half the time.
What it actually measures
MSAVI2 replaces SAVI's fixed L with a self-adjusting function of NIR and red, solved so the soil correction strengthens where vegetation is sparse and relaxes where it is dense. The square-root form looks intimidating, but it is simply the closed-form solution that makes the adjustment adaptive rather than constant.
The math behind it
where NIR is near-infrared reflectance, Red is red reflectance.
How to interpret the output
Interpreted like SAVI; higher = denser green vegetation. Best where canopy cover changes through a season and a fixed L would be wrong.
The bands it needs
| Sentinel-2 | NIR B8 · Red B4 |
| Landsat 8/9 | NIR B5 · Red B4 |
Real-world use
It shines in time-series work across a growing season, where canopy cover swings from bare soil to full closure and any fixed L would be wrong at one end. Precision-agriculture emergence mapping is a classic use.
Watch out for
MSAVI2 still assumes a single soil line and still tops out at very high biomass — it fixes SAVI's tuning problem, not saturation. For dense evergreen canopy, reach for EVI instead.
Compute MSAVI2 on your own study area
Skip the code. Draw or upload a boundary and Spatial Research Suite runs MSAVI2 on live Sentinel-2 or Landsat imagery — with cloud masking, exports and citations built in.
Run this analysis in GISforus →Frequently asked
Why use MSAVI2 instead of SAVI?
MSAVI2 removes the need to choose the soil factor L by computing it adaptively from the scene, which matters when canopy density varies within an image or over a season.
Does MSAVI2 fix NDVI saturation?
No — it addresses soil background, not saturation. Use EVI for very dense, high-biomass canopy.