How do you read an NDVI map?
NDVI compares the red and near infrared light reflected by a surface. Higher values often accompany active green vegetation, but they do not by themselves identify a species, measure crop yield, diagnose plant health or prove drought. Read the surface reflectance product, cloud mask, acquisition date and pixel size together before interpreting a map.
Calculation and map layers
1. How do two bands produce the index? Normalized Difference Vegetation Index is calculated as near infrared minus red, divided by near infrared plus red. Chlorophyll in green leaves absorbs much red light, while the internal leaf structure reflects more near infrared light. The resulting contrast can raise the ratio. Use an appropriately corrected surface reflectance product rather than assuming that raw image brightness is directly comparable across dates. Check the product's missing data rule and pixels whose denominator is extremely small; a number without a valid observation is not a vegetation measurement. 2. What do colors and values mean? The theoretical range runs from negative one to positive one. A value around zero or below may represent water, cloud, bare rock or built surfaces; a single threshold cannot separate all these classes reliably. Medium to high positive values often indicate more green plant signal. A cartographer chooses the color palette independently of the data, so a vividly green patch has no numeric meaning until you read the legend. When comparing two dates, keep the same classification ranges and color ramp, or a visual shift may be a display choice rather than physical change. 3. What lies inside one pixel? A Landsat pixel can mix a narrow road, field edge, shadow and plants. Its ratio does not give the condition of every plot within the footprint. Compare sample pixels with higher resolution imagery, a land cover layer and, where possible, field observations. Pixel size matters especially for small holdings and individual urban trees. An average over a district can hide strong differences among fields, slopes and irrigation systems, even when its arithmetic calculation is correct.
Time series and common misreadings
4. How can seasonal change be separated from loss? A difference between a spring image and a late summer image may reflect ordinary plant growth and senescence rather than damage. Compare similar stages of the same vegetation type, preferably over several years and under comparable observing conditions. Irrigated farmland and an unirrigated hillside can respond differently to one weather event. Harvesting, crop rotation and land conversion can also alter the NDVI trace. Never call a fall in greenness drought without dates, land use context and other evidence such as rainfall and soil moisture. 5. How do cloud and shade affect a map? Clouds, cloud shadows, snow and haze can hide the true surface signal. The USGS Landsat products include quality assessment bands with flags for these pixel conditions. Read the bit definitions for the particular sensor and product collection, then mask unreliable observations. Do not color gaps left by masking as though they were actual vegetation loss. If the valid area differs markedly between dates, even a correctly computed regional mean may be misleading because it describes different sampled ground. 6. What question can the map answer? NDVI can begin an investigation of seasonal greening in a wetland or spatial contrasts within a field experiment. Species identification, biomass and harvested yield need extra measurements and calibration. The index can approach saturation in very dense leaf cover, while bare soil contributes substantially to sparsely vegetated pixels. State the sensor, processing level, acquisition dates, mask, pixel size, thresholds and uncertainty alongside any conclusion. A useful map is a documented measurement with limitations, not merely an attractive green image.