NDVI
NDVI (Normalized Difference Vegetation Index) is a computationally generated vegetation index map that makes it convenient to assess vegetation density and overall condition. It can also be referred to more simply as a vegetation greenness map. Chlorophyll in plants reflects near-infrared radiation (NIR – Near Infrared) and green light (G – Green) effectively, while absorbing visible red light (R – Red). This map layer allows for an evaluation of the difference between reflected near-infrared and red radiation.
The calculation is performed using the following formula:
NDVI is normalized, meaning its values are scaled to a specific range, typically from –1 to +1. The higher the NDVI value of an area, the greener it appears on the map. Consequently, the vegetation in such areas is also denser and healthier (Figure 1). Vegetation in very good or excellent condition (dense, lush, green) may not distinguish well from each other on the NDVI layer, as the green tone in the gradient saturates beyond a certain level. However, areas with sparse greenery, uneven crops, plowed/damaged soil, or other conditions requiring closer attention are particularly easy to identify.
Image 1. Vegetation index image of summer greenery with vegetation over 10m in Lääne-Nigula municipality (black areas correspond to vegetation under 10m).
Thus, a green tone indicates that the vegetation is likely healthy and in good condition. When the vegetation is sparser, more red light is reflected, and the proportion of near-infrared radiation decreases. In areas with low, sparse, damaged, or absent vegetation, the vegetation index map displays tones ranging from yellow to red (Figure 2).
In general, red tones indicate plowed land, absence of vegetation, or areas affected by erosion. Yellow tones signify vegetation that is not in optimal condition, such as sparse or partially exposed soil. Low NDVI values may result from dryness, pests, or other factors that reduce greenness. However, it is important to remember that low NDVI values can also indicate other types of areas, such as artificial surfaces, water, or ice. Therefore, it is necessary to examine the area more closely to determine what the low NDVI value refers to in a specific case.
Figure 2. Agricultural areas combined with an open area mask and PRIA field boundaries applied (reference to Satiladu for detailed exploration).
The NDVI data layer is smoothed for more pleasant viewing, but if desired, the user can disable the smoothing function in Satiladu by clicking the button with pixel squares located in the top-right corner (Figure 3). This will display the original pixels. Viewing original pixels may be more appropriate for various scientific or analytical tasks where maintaining pixel-level accuracy is critical to avoid the uniformity caused by smoothing, which might impact the precision of the analysis.
For example, if a person is in nature and uses mobile location tracking (enabling location detection in Satiladu), they can use the aforementioned button to make the image "pixelated." This allows them to see the pixel value of the location they are currently at or navigate to a pixel square of interest, observing different landscape units or vegetation conditions within their movement area. NDVI pixel values can be viewed in Satilaos from the bottom-right corner next to the coordinates field (Figure 4). NDVI pixel values are accessible across all map layers by moving the cursor.
Figure 3. In Satiladu, NDVI data can be viewed as a smoothed image or as pixel squares.
Figure 4. A visual representation of the range of values associated with different shades of land cover types.
NDVI is primarily designed for vegetation analysis, but it can also be used across various other fields. Negative NDVI values are likely to indicate the presence of a water body. These are represented in deep blue tones on the Satilaos NDVI map layer (Figure 5). When observing water bodies, it is recommended to disable the Hybrid layer using the button in the bottom-left corner.
Although NDVI cannot directly assess the quality of water bodies (e.g., pollution or temperature), it can provide clues about the presence of vegetation in the water, such as algae or other aquatic plants.
Figure 5. Visual of the Narva reservoir through the vegetation index map layer.
When using the NDVI map layer, it is important to recognize that NDVI is sensitive to various atmospheric effects. Clouds, for instance, reflect more red light than near-infrared radiation, which is why they appear red on the map (Figure 6). Some thinner clouds may also appear yellow. Therefore, it is advisable to use an RGB photo image as a reference alongside the NDVI map layer to ensure that features, such as cloud cover, are correctly identified and not mistaken for vegetation anomalies.
Figure 6. Clouds appear in yellow, red, and black tones, depending on their density.
NDVI is most effective for monitoring the green period, as it is specifically designed to evaluate vegetation health and vitality, which are best observed during warm, green growing seasons. In winter, when vegetation is less active or completely dormant, NDVI data becomes harder to interpret because snow, ice, and frozen soil can distort the results. Consequently, NDVI is not the optimal tool for analyzing the condition of winter vegetation. However, for those who seek and observe closely, intriguing insights can still be discovered.
In agricultural areas, including fields shaded by forest edges, spring snowmelt may be uneven. In some regions, ice and snow can persist longer into the spring, impacting the soil and vegetation. Depending on the situation, the lingering snow may have either a negative or positive effect. Snow-covered and icy puddles are displayed on the NDVI map layer in bluish-black tones (Figure 7).
Figure 7. Comparison of February NDVI (right) and RGB (left). An open area mask has been applied to highlight the fields.
The image below (Figure 8) provides an example of how winter barley survived the winter and spring cold spells specifically under a snow patch. The use of NDVI during the winter is rather niche and requires familiarity with the area being studied. This example illustrates how retrospective NDVI analysis can provide indications of microclimatic conditions or factors, such as snow patches, that may have influenced the growth and development of crops. However, final conclusions largely depend on additional studies and the specifics of the local environment.
Figure 8. NDVI map layer on the left during the cold period (2024-03-04) and on the right during the summer period (2024-06-24), indicating that the long-lasting spring snow cover had a positive effect on the crops.