BAE 2023

Color Measurement Lab: Use of a spectrometer to determine NDVI

Background:

The use of color to determine plant vigor is an old idea. Almost anyone can deduce that the “greenness” of a plant is a good indicator of plant health or plant stress. Traditionally in large-production agriculture, a management decision, such as top-dressing fertilizer, has been made from viewing the overall health of a field, treating the field as one unit, and applying the average amount of fertilizer needed. However, high chemical costs and environmental concerns encourage producers to find a more selective method of fertilizer application. Additionally, a standardized measurement of plant vigor can help scientists determine plant biomass and chlorophyll concentration, which can be used in environmental models to predict water usage, evapotransportation, photosynthesis, nutrient use, etc.

The cones in our eyes contain three types of light receptors, each identifying light from a specific region of the visible spectrum (380 to 780 nm). We integrate the electric impulses from the three types of receptors to interpret intensities and colors. However, most electronic light detectors are sensitive to all light within a region of the electromagnetic spectrum. Thus, for a machine to detect color, the color must be isolated, the intensities detected, and the data interpreted.

The reason a healthy plant appears green is because, in the visible spectrum, red and blue light are absorbed and green light is reflected. Thus, looking at the spectrum of a healthy plant, we would expect to see a low reflectance of red light (670 nm). In addition, the structures of healthy plant cells reflect near-infrared light (NIR). Thus, we would expect a healthy plant to have high reflectance in the NIR range (780 nm).

One method for machines to determine the vigor of a plant canopy relies on the reflectance of particular wavelengths of light. The detected reflectance is used to establish a vegetation index, which indicates the vigor of the plant. The normalized difference vegetative index (NDVI) is one such index that relates the reflectance of red and near-infrared light.

This laboratory will feature the use of a spectrometer to measure light ranging from 300 nm to 1000 nm. The light emitted from fluorescent and incandescent bulbs will be viewed and color differences of different surfaces will be identified. The intensities of red (670 ± 10 nm) and near-infrared (780 ± 10 nm) light will be used to calculate NDVI.

Materials:

  1. Field Spec Spectrometer
  2. Laptop Computer
  3. Barium sulfate painted plate
  4. Various surfaces: fabric, soil, plant material

Before you begin, you should familiarize yourself with the equipment. The spectrometer you are using relies on a fiber optic cable to capture light and transmit it through a refracting gradient and onto several detectors. The spectrometer is interfaced with a laptop computer, which is used to record the measured data.

Procedure:

  1. Turn on the spectrometer and the laptop computer.
  2. Insert the fiber optic cable into the supporting holder
  3. Take a white and dark current reading with the cable and holder placed directly on the calibration target.
  4. Notice that the Y-axis has now changed to represent reflectance
  5. Without changing lighting conditions, record spectra of the various surfaces
  6. Transfer all of the saved data into a text file use the FieldSpec software.

Data Processing:

Plot and compare the spectra from the various surfaces you considered. On these graphs, they y-axis should have units of reflectance.

For each of the items considered, compute the average reflectance value ranging from 660 nm to 680 nm. This number is the red reflectance value (IRED). Compute the average of the reflectance values ranging from 770 nm to 790 nm. This is the near-infrared reflectance (INIR). Now compute NDVI for all of your samples. Create a table that displays your data and observations.

In your report, comment on the use of NDVI to detect healthy plants and unhealthy plants and why it is important to accurately determine plant health and growth status in modern agriculture.