What is MSAVI2? (Modified Soil-Adjusted Vegetation Index)
What is MSAVI2?
MSAVI2 is the self-adjusting successor to SAVI. Where SAVI requires you to pick a fixed soil-brightness factor L (usually 0.5) and accept the compromise, MSAVI2 computes the ideal L from the data itself, pixel by pixel. The result is a soil-corrected vegetation index that adapts to whatever density it encounters — making it the best choice for heterogeneous fields, very sparse vegetation, and early growth monitoring. If SAVI is the early-season specialist with a fixed tool, MSAVI2 is the same specialist with a self-sharpening one.
How it’s calculated
MSAVI2 uses the same NIR and Red bands as NDVI and SAVI, but replaces the fixed L with an in-formula iterative solution that derives L from the vegetation density in each pixel:
MSAVI2 = (2 · NIR + 1 − √((2·NIR + 1)² − 8·(NIR − Red))) / 2On Sentinel-2 these are Band 8 (NIR) and Band 4 (Red). The formula starts from the SAVI equation and solves for the L that minimizes soil background noise given the actual reflectance values. The square-root term is that self-adjustment in action — it produces a different effective L for every pixel.
Unlike SAVI, MSAVI2 has no free parameter to set and no DN-scaling pitfall (the ratio structure is self-balancing). This makes it both more accurate and simpler to implement correctly.
Typical value ranges
MSAVI2 occupies the same −1 to +1 range as NDVI and SAVI. For sparse canopies, it sits above NDVI (removing the soil penalty) and close to SAVI with L = 0.5.
| MSAVI2 range | Meaning | Typical for |
|---|---|---|
| 0.6 – 0.9 | Dense, healthy vegetation | Peak-season closed canopy |
| 0.4 – 0.6 | Moderate vegetation | Developing crops, healthy pasture |
| 0.2 – 0.4 | Sparse but viable | Young crops, arid rangeland — MSAVI2’s sweet spot |
| 0.0 – 0.2 | Very sparse / bare | Newly emerged crop, dry bare ground |
| Below 0 | Water, snow, cloud | Open water, snow cover |
When to use it
Use MSAVI2 for sparse vegetation (young crops, arid rangeland, shrubland), heterogeneous fields with mixed crop stages, and areas with variable soil brightness. The more variability you have in vegetation density or soil brightness, the more MSAVI2 pulls ahead of SAVI. Over a season, MSAVI2 follows the same growth curve as NDVI but reads higher in the early stages — that gap is the soil-brightness correction. For a heterogeneous field with patchy establishment, MSAVI2’s spatial map shows less soil-driven noise than NDVI, making stressed zones easier to spot.
Once the canopy closes and soil is no longer visible, MSAVI2 converges with NDVI — switch back to NDVI there. The square-root term can fail on bad data (negative under the root from sensor noise), producing NaN — robust implementations clamp or filter these cases.
Comparison with other indices
MSAVI2 is NDVI with automatic per-pixel soil correction. SAVI is the fixed-L predecessor — MSAVI2 is more accurate whenever vegetation density varies within a field. NDVI remains the universal default for closed-canopy monitoring. For high-biomass saturation issues, EVI is the right choice. MSAVI2’s self-balancing formula also avoids SAVI and EVI’s DN-scaling pitfalls entirely, making it both more robust and simpler than either.
Try it free: Get MSAVI2 values for your field — free, no signup. Works on any land worldwide. Check my field’s establishment →
Frequently asked questions
What is the difference between MSAVI2 and NDVI?
MSAVI2 is NDVI with an automatic soil-brightness correction. NDVI assumes only vegetation affects the reflectance; MSAVI2 self-computes a per-pixel soil factor so that bare soil showing through a sparse canopy does not drag the index down. The two converge once the canopy closes; MSAVI2 matters most in early growth and heterogeneous fields.
What is the difference between MSAVI2 and SAVI?
SAVI uses a fixed soil-adjustment factor L = 0.5. MSAVI2 computes L automatically from the data using an iterative formula, so the correction adapts to each pixel’s actual vegetation density. For fields with variable soil brightness or mixed crop density, MSAVI2 is more accurate; for uniform conditions, SAVI is simpler and equally valid.
When should I use MSAVI2?
Use MSAVI2 for sparse vegetation (young crops, arid rangeland, shrubland), heterogeneous fields with mixed crop stages, and areas with variable soil brightness. Once the canopy closes and soil is no longer visible, switch to NDVI — the correction adds no value and the indices converge.
Does MSAVI2 require a DN-scaling fix like SAVI?
No — this is one of MSAVI2’s advantages. Because its formula is a ratio-like structure that self-balances, it does not have the DN-scaling pitfall that SAVI and EVI have (where the L factor must be multiplied by 10000 in Sentinel-2 DN space). MSAVI2 is simpler to implement correctly.