Remote Sensing of Vegetation

The brightness or reflectance of vegetation varies across the electromagnetic spectrum. We see healthy vegetation as predominately green with our eyes and in color photos (Figure 1a below) because the pigments (e.g.: chlorophyll) in plants are absorbing blue and red radiation to carry on photosynthesis. When we look in the near-infrared region using color-infrared photography (Figure 1b below), which is just beyond what we can see with our eyes, actively growing plants are highly reflective because of the multiple scattering that takes place between the spongy-mesophyll cells of the plant.

a. Natural Color	b. Color-infrared

Figure 1. Airborne digital color and color-infrared images of a public building and residential neighborhood in Hartford, CN. The data were obtained at a spatial resolution of 0.5 x 0.5-m. a) Natural color image (RGB = red, green, blue). b) Color-infrared image (RGB = near-infrared, red, and green) (courtesy Positive Systems, Inc.) (Jensen, John R. Remote Sensing of the Environment: An Earth Resource Perspective. Prentice-Hall, New Jersey.)

Actively growing plants therefore show a strong contrast between strong absorption in the red and high reflectance in the near-infrared regions of the spectrum. The amount of absorption in the red and reflectance in the near-infrared varies with both the type of vegetation and the vigor of the plants. The natural change of a green leaf as it senesces is shown in Figure 2. We see that the healthy green leaf has very low reflectance values in the red (600-700 nm) due to chlorophyll absorption and very high reflectance values in the near-infrared (700-1000 nm). However, as the plant begins to senesce, reflectance begins to decrease in the near-infrared and increase in the red.

Figure 2. a) Healthy green Sweetgum leaf obtained from a tree on November 11, 1998. b-c) Senescing yellow and red Sweetgum leaves obtained from the tree. d) Senesced Sweetgum leaf that was on the ground. e) Reflectance measurements of the four leaves. (Jensen, John R. Remote Sensing of the Environment: An Earth Resource Perspective. Prentice-Hall, New Jersey.)

The decline in reflectance in the near-infrared is due to the spongy-mesophyll layer collapsing as the leaf comes under stress and the increase in reflectance in the red is caused by the die-off of chlorophyll and therefore a decrease in absorption.

We can therefore use this information to evaluate vegetation type, condition, or density. One way of characterizing this relationship with a single variable is by dividing the near-infrared reflectance by the red reflectance (NIR/Red). The larger this ratio, the more photosynthetically active vegetation is present; the lower the ratio, the less photosynthetically active vegetation present. With aircraft and satellite digital sensors we acquire brightness or reflectance data in separate regions of the electromagnetic spectrum. This allows us to create a NIR/Red ratio image simply by dividing the NIR image by the Red image. Because these ratio values will vary considerably from one region to another, a way of normalizing the ratio was established called the Normalized Difference Vegetation Index or NDVI. NDVI = (NIR – Red) / (NIR + Red). The NDVI image of Arizona (Figure 3) shows the range of NDVI across the state.

Figure 3. NDVI image of Arizona from the MODIS sensor acquired on March 14, 2002. The image has been color coded so that areas of high NDVI appear in shades of green and areas of low NDVI appear in shades of tan and brown.

Figure 4: Generalized relationship between NDVI and LAI for a range of soil and vegetation spectral properties, including soil, snow, and herbaceous leaf optical signatures. Note that the same color scheme is used in both Figures 3 and 4, to allow for an estimate of the LAI value.

The NDVI has been related to biophysical parameters like vegetation density, percent green vegetation cover, biomass, fraction of absorbed photosynthetically active radiation (fAPAR) and leaf area index (LAI). The relationships of NDVI and a biophysical parameter are often site specific and vary with the soil and vegetation optical properties (e.g. leaf spectra, leaf angle distribution, plant architecture). The position of the sensor and sun also affect the relationship between the NDVI and a biophysical parameter. Figure 4 shows a generalized example of how the NDVI is related to the leaf area index (LAI). If there is no vegetation, the NDVI values are lowest for snow covered areas and bright soils, and are generally higher for dark soils with high organic matter contents. Figure 4 indicates that multiple NDVI values have multiple LAI values associated with them, primarily due to variable vegetation and soil types and season (phenology). The relationships between NDVI and LAI are often less variable if developed for one particular vegetation and soil type.

If we plot the red against the near-infrared reflectance for an entire image (Figure 5) we generally see a triangular shaped distribution as shown in the shaded area. The soil line represents areas on the ground that do not have significant vegetation cover but vary in reflectance based upon whether the soil is wet or dry. Areas where vegetation is present will be found somewhere above the soil line.

Figure 5. a) Distribution of reflectance values in a remote sensing image in the red and near-infrared regions of the electromagnetic spectrum are found in the gray shaded area. The greater the amount of photosynthetically active vegetation present, the greater the near-infrared reflectance and the lower the red reflectance. This condition moves a pixel’s spectral location in a perpendicular direction away from the soil line. b) The migration of a single vegetated agricultural pixel in red and near-infrared multispectral space during a growing season is shown. After the crop emerges, it departs from the soil line, eventually reaching complete canopy closure. After harvesting, the pixel will be found on the soil line, but perhaps in a drier soil condition. (Jensen, John R. Remote Sensing of the Environment: An Earth Resource Perspective. Prentice-Hall, New Jersey: pg. 343.)

Our ability to map and monitor vegetation condition continues to improve as we make operational new satellite sensors that increase our ability to image smaller areas on the ground (spatial resolution) and in more regions of the electromagnetic spectrum (spectral resolution).