Step 1: Recall the limitations NDVI has that later indices try to fix.
The Normalized Difference Vegetation Index, \(NDVI = \dfrac{NIR-Red}{NIR+Red}\), is simple to compute but suffers from three known problems: it is sensitive to atmospheric effects such as aerosol and water vapour scattering, it is affected by the reflectance of exposed soil in sparsely vegetated areas, and it saturates at high biomass or leaf area index, so it cannot distinguish moderately from very densely vegetated canopies.
Step 2: Recall how SAVI addresses only one of these problems.
The Soil-Adjusted Vegetation Index introduces a soil brightness correction factor \(L\), \(SAVI = \dfrac{NIR-Red}{NIR+Red+L}(1+L)\), which reduces the influence of soil background reflectance, but it has no correction for atmospheric scattering and still saturates at high canopy density, so it solves only the soil problem.
Step 3: Recall how EVI is built to fix all three problems together. \[ EVI = G\cdot\dfrac{NIR-Red}{NIR+C_1\,Red-C_2\,Blue+L} \] The blue band term \(C_2\,Blue\) is an aerosol resistance term that corrects for atmospheric scattering, the canopy background factor \(L\) (carried over from SAVI) reduces soil background influence, and the gain factor \(G\) with the modified denominator improves sensitivity in high-biomass regions, delaying saturation compared to NDVI. EVI is therefore specifically engineered to jointly correct for atmosphere, soil background, and saturation.
Step 4: Eliminate the remaining options.
NDVI (B) has none of these corrections built in; it is the baseline index these problems were identified from. LAI (D) is not a spectral vegetation index at all, it is a biophysical canopy parameter (leaf area per ground area) that is often estimated from vegetation indices, not an index that corrects for these three effects.
Step 5: Conclusion.
Only EVI is purpose-built to jointly handle atmospheric effect, soil background effect, and saturation at high vegetation density.\[ \boxed{\text{EVI (option A)}} \]