Lesson 4 of 6 · Earth observation
Measure vegetation before naming its cause
9 minute lesson · 3 checks along the way
Your goal for this lessonUse a spectral index as evidence while distinguishing cover, use, and ecological change.
Words with a dotted underline open a short definition right here in the lesson.
What a vegetation index can summarize
Plants interact with wavelengths differently. Leaf pigments absorb much visible red light, while internal leaf structure strongly affects scattering. The normalized difference vegetation index, , compares those two bands: near-infrared minus red, divided by near-infrared plus red. It is a dimensionless spectral ratio. It can track patterns associated with vegetation, but it is not itself a count of trees, a crop yield, or a carbon inventory.
Use properly decoded, comparable reflectance values rather than assuming stored digital integers are already physical reflectances. Product metadata can specify both a multiplicative scale and an additive . A common multiplier cancels in the ratio, but a common additive offset changes its denominator and does not generally cancel. Quality masks, valid ranges, and a nonzero denominator must also be checked before calculating or interpreting an index.
Distinguish land cover from land use
describes physical surface classes such as trees, water, or buildings. concerns activities and purposes, such as conservation, grazing, or cultivation. A satellite can reveal patterns that help investigate use, but similar vegetation can occur under different ownership and management. Seasonal harvest, drought, fire, flooding, and permanent clearing can all change a vegetation signal through different processes.
Comparable observations therefore need suitable dates, illumination, atmospheric correction, and spatial alignment. A leaf-off winter scene should not be mistaken for the same ecological state as a summer scene. Soil background and mixed influence the ratio, and dense canopies can make NDVI less responsive to additional vegetation. Independent field observations and other spectral bands help test what a change means in a particular landscape.
Check 1 of 3
10 pointsUse the reading above. You can try again without losing points.
Work through a decision
Calculate a change without overstating it
For an invented parcel, let near-infrared reflectance be 0.60 and red reflectance 0.20. NDVI is (0.60−0.20)/(0.60+0.20)=0.50. Later, comparable reflectances are 0.40 and 0.20, giving 0.20/0.60, or about 0.33. The index decreased by about 0.17. That is a change in a ratio, not direct evidence that 17 percent of the trees disappeared.
Suppose a separate, validated map identifies twelve complete 30-by-30-meter cells as changing land-cover class. Their mapped area is 12×900=10,800 square meters, or 1.08 hectares. This area calculation uses the classified cells, not the NDVI difference. Boundary mixtures and classification errors affect the actual changed area, so the estimate needs and an explicit class definition rather than a bare number.
An index change keeps its own units
Original hypothetical comparable reflectances with a shared zero-to-one axis. The formula applies to decoded reflectances; its dimensionless result needs additional evidence before an ecological conversion.
Open full-size diagram in a new tab: An index change keeps its own unitsCheck 2 of 3
10 pointsUse the reading above. You can try again without losing points.
Test explanations with more than two pictures
Ask whether the change persists through comparable later seasons and whether nearby reference parcels behave similarly. A temporary index decrease followed by recovery could fit a different explanation from a new road network and persistent conversion to buildings. Neither pattern alone establishes motive or legal status. Labels should describe the evidence actually observed before they assign a human action or ecological outcome.
The Landsat record and Sentinel-2 observations support such repeated comparisons when their processing is made compatible. A reproducible result preserves the selected scenes, masks, region, index formula, and classification method. It also records which independent observations were used to validate the conclusion. That record lets another reader distinguish a real landscape signal from a seasonal, instrumental, or analytical difference.
Check 3 of 3
10 pointsUse the reading above. You can try again without losing points.
Put it together
Your lesson progress
Each check connects to a part of the story. Complete all 3 to earn your 25-point lesson bonus.
Go to the source
Explore the evidence behind this lesson. These optional readings and videos open in a new tab.