What the NDVI index is and how it makes a farmer's life easier

Reading time — 15 min
Here is a simple explanation of what the NDVI vegetation index is and how to use it for field analysis.
Variable-rate sunflower seeding_Cover_OneSoil Blog

How light helps in understanding plant health

Put it simply, the NDVI (Normalized difference vegetation index) is an indicator of a plant's health based on how a plant reflects different light waves.

For example, for the human eye, a plant is green because the chlorophyll pigment in it reflects green waves. Chlorophyll also absorbs red waves and so photosynthesis occurs. In other words, a plant grows and develops. The cell structure of a plant reflects the near-infrared waves. So a healthy plant, the one with a lot of chlorophyll and good cell structure, actively absorbs red light and reflects near infrared. And exactly the opposite happens to a diseased plant. Healthy plant actively absorbs red light and reflects near infrared.

NDVI formula

To understand the state of a plant's health, we need to compare the values of absorption and reflection of red and infrared light. This is what the NDVI index is used for. It is calculated using the following formula (side picture).
NDVI formula_OneSoil blog
NDVI formula
The index became "normalized" in 1973 when a group of scientists from the Texas A&M University began to calculate not the ratio of infrared light to red, but the ratio of their difference to a total — this is actually the modern formula. It allowed for the ranging of all the index values from -1 to 1, to normalize them for ease of comparison. NDVI values between -1 to 0 correspond to surfaces like snow, water, sand, stones and infrastructure objects, like roads and houses. NDVI values for plants range from 0 to 1.

NDVI interpretation at different stages of the season

It is important to understand that the NDVI is an indicator of the plant’s health but it says nothing about the cause of a particular condition. The vegetation index is rather a hint at what is currently happening on the field. Let’s consider three scenarios of NDVI usage for field analysis: at the beginning, in the middle, and at end of the growing season.

At the beginning of the season, the NDVI index helps to understand how the plant has survived through the winter.
  • 1
    If the NDVI is lower than 0.15, most probably all the plants died in this part of the field. Typically, these figures correspond to plowed soil without any vegetation.
  • 2
    0.15−0.2 is also a low value. This may indicate that plants started wintering in the early phenological phase, before tillering.
  • 3
    0.2−0.3 is a relatively good value. Probably, the plants entered the tillering stage and have resumed vegetation.
  • 4
    0.3−0.5 is a good value. Nevertheless, one should keep in mind that high NDVI values can indicate that plants wintered at a late phenological stage. If the satellite image was taken before the resuming of vegetation, then it is necessary to analyze the zone after the resuming of the vegetation also.
  • 5
    Above 0.5 is an abnormal value for the post-wintering period. It is better to check this field zone yourself.
To sum up, if you see abnormal NDVI values (those that are very different from the average values for the field), you need to check this field area. You can see the NDVI index for your fields, monitor them and leave notes for free in our OneSoil Scouting Android app or on the OneSoil web platform. When the weather is cloudless, images are updated every 3−5 days.
NDVI vs real field photo_OneSoil blog
That’s how the NDVI index on the OneSoil web platform correlates
with the real state of plants in the field
In the middle of the season, the NDVI index helps to understand how plants grow and develop. If the index values are medium to high (0.5−0.85), most likely there are no major issues at this part of the field. If the index is low, probably there are specific issues, like a lack of moisture or nutrients. It is better to check this part of the field yourself.

Using the NDVI index, we create maps for variable-rate application of nitrogen. We detect areas of low, medium and high vegetation indices, and then a farmer himself can define the rate of fertilizer required. In our experience, the optimal nitrogen application scheme is as follows
  • 1
    If the vegetation index is high in the area, the fertilizer dose should be reduced to 10−30% of the average rate.
  • 2
    If the vegetation index is average, the fertilizer dose should be increased up to 20−25% of the average rate.
  • 3
    If the vegetation index is low, you need to determine the cause of it first.
We also use the NDVI index to restore the crop yield for a field. Using this information, we create maps for variable-rate application of phosphate and potassium fertilizers.

At the end of the season, the NDVI index helps to determine which fields are ready for harvesting — the lower the index, the closer the part of the field is to the ripening stage. The optimal index value, in this case, would be lower than 0.25.

NDVI in remote sensing: pros and cons

There are many vegetation indices, and most of them are similar to each other. But the NDVI is the most popular and widespread one, and it also has one important advantage: the high resolution of images with data from the Sentinel-2 satellite. In such cases, channels with a resolution of 10 meters are used (1 pixel is 10 by 10 meters) to calculate the NDVI index. Those indices that use additional light channels, mostly red age, have a resolution of 20 meters (1 pixel is 20 by 20 meters). In OneSoil, we use only space images from the Sentinel-2 satellite.

One of the drawbacks of the NDVI is that the index loses sensitivity when a plant reaches a certain development threshold. In other words, if a plant develops very actively, it becomes impossible to distinguish an abnormally green plant from a "normally" green one. The NDVI index also depends on the weather: if clouds are overhanging a field for a long time, the satellite image will be unclear. However, this is true for any index, not just the NDVI. To solve this issue, we at OneSoil use the radar data from the satellite Sentintel-1 for some tasks, such as crop recognition.

3-click solution for perfect variable-rate trials
Related Articles
If you want to discuss this article and share your own experience, please join our Telegram or WhatsApp!
OneSoil
People, fields, and technology
Get all new articles straight to your inbox! You can read them even in the field.