Remote Sensing Capabilities
Over time and through many scientific studies, remote sensing experts have come to understand how the measured reflectance of crops at two or more wavelengths reveal important vegetation information. These reflectance measurements are used to calculate a multitude of various vegetation indices (VIs). The math is easy – most include subtracting or dividing reflectance values at multiple wavelengths. You can throw in a few first derivatives if you enjoy math but – that's not required as PIC will do the "heavy lifting" for you and present VIs with important characteristics of your crops shown overlaid a map of your fields.
The high spectral and spatial resolution PIC sensor suite flying over your crops generates whole-field metrics that can be reviewed by you block by block, row by row or vine by vine. The PIC sensor suite will accurately sample your entire crop rather than applying scant/sparse ground measurements to a large area. The vine by vine data we provide across your entire vineyard will provide information you can use throughout the growing season to maximize your efficiency.
In this section we will introduce you to the agricultural basis of these VIs and explain uses and limitations of a variety of VIs.
Currently there are more than 150 existing VIs, with additional indices emerging as sensors advance and provide new information. In collaboration with our University and government-based scientific collaboration team, PIC will provide you the most useful and relevant VIs to help you make informed decisions and manage your crop throughout the growing season.
Red Edge & Vegetation Indices
Most vegetation indices are derived from reflectance values in the visible (to the human eye) waveband (400 – 700 nanometers). For reference, blue light is approximately 475nm, green is approximately 510nm and red is approximately 650nm.
Figure 1 Visible wavelengths
The visible waveband (Figure 1) is used by plants for photosynthesis. Shorter wavelengths (<400 nm – aka ultraviolet) tend to be so energetic that they can be damaging to cells and tissues (e.g. sunburn in humans).
Longer wavelengths (e.g. infrared) do not carry enough energy to power photosynthesis; however, they can be used to assess changes in nitrogen content/uptake, canopy/leaf water content and onset of senescence. These IR vegetation indices require sensors sensitive to near infrared (700 – 1,300 nm) and/or shortwave (1,300 – 2,500 nm) IR light. PIC will be providing IR vegetation indices in the near future.
Figure 2 Red edge in vegetation reflectance curve
Before discussing any VI in detail it would be useful to understand the "red edge" – a reflectance feature (Fig. 2) unique to vegetation that is precisely measured by our high spectral resolution sensor and used in calculations of most VIs.
Chlorophyll, the most abundant plant pigment, absorbs red and blue light. Other pigments such as carotenes and xanthophylls absorb some green light and pass it on to the photosynthetic process, but enough of the green wavelengths are reflected to give leaves their characteristic color. In autumn, when chlorophyll is degraded, carotenes & xanthophylls remain in the leaf producing red, yellow and orange leaves. Red Edge refers to the region of rapid change in reflectance of vegetation in the near infrared. As noted above, chlorophyll, strongly absorbs visible light (from 0.4 to 0.7 µm in Fig. 2) for use in photosynthesis. The cell structure of the leaves, on the other hand, strongly reflects near-infrared light (from 0.7 to 1.1 µm). The resulting rapid increase in reflectance, from 5% in the visible to 50% in the NIR) is referred to as the red edge as shown in Figure 2.
The red edge accounts for the brightness of foliage in infrared photography and is extensively utilized in calculations of vegetation indices.
NDVI & Broadband VIs
There are two general classes of vegetation indices: 1) those that can be calculated from two or three waveband sensors – called broadband VIs and 2) those that require the spectral resolution obtained by advanced hyperspectral sensors (such as used by PIC) – called narrow band VIs.
The Normalized Difference Vegetation Index (NDVI) is a broadband VI. It is the first ever and most commonly used vegetation index (VI). The earliest reported use of NDVI was in 1973.
NDVI was a product of the Landsat 1 satellite launched July, 1972 by NASA to examine Spring, Summer and Fall changes in vegetation throughout the Great Plains region of the central U.S.
Landsat's ability to measure the characteristics of vegetation from satellite spectral signals was confounded by differences in solar zenith angle (how high or low the sun is in the sky). NASA studied solutions and developed the ratio of the difference of the red and infrared radiances over their sum (i.e NDVI) as a means to adjust for or "normalize" the effects of the solar zenith angle (Figure 3).
NDVI was one of the most successful of many attempts to simply and quickly identify vegetated areas and their "condition" It remains the most well-known and (over)used index to detect live green plant canopies in multispectral remote sensing data. Once the utility to detect vegetation had been demonstrated, users used NDVI to quantify the photosynthetic capacity of plant canopies.
NDVI values range from -1 to 1. Negative values of NDVI correspond to areas of water. Values close to zero generally correspond to barren areas. Low, positive values represent shrub, grassland or stressed vegetation while high values indicate healthy vegetation.
Broadband VIs are the simplest measure of the general quantity and vigor of green vegetation. Broadband greenness VIs compare reflectance measurements from the reflectance peak of vegetation in the near-infrared range to reflectance taken in the red wavelength range where chlorophyll absorbs photons to store into energy through photosynthesis. As such they are all "red edge" indices.
NDVI and other broadband indices are sensitive to the combined effects of chlorophyll concentration, canopy leaf area, foliage clumping, and canopy architecture. These VIs are designed to provide a measure of the overall amount and quality of photosynthetic material in vegetation – essential for understanding the state of vegetation for any purpose.
Increases in chlorophyll concentration or leaf area or decreases in foliage clumping (i.e. changes in canopy architecture) will increase broadband VI values. They do not provide quantitative information on any single environmental factor (water, fertilizer, etc.) contributing to canopy vigor.
A number of issues may seriously limit the actual usefulness of NDVI and other broadband VIs including: atmospheric effects, shadows, wet soils, seasonal and daily sun angle differences, etc. Derivatives of and alternatives to NDVI have been developed by University researchers, USDA and foreign agricultural institutes. These alternative VIs include corrections for one or more of these issues. Also, NDVI has tended to be over-used in applications for which it was never designed. For these reasons, NDVI should be used with great caution and supplemented with other broadband and/or narrowband indices.
PIC provides multiple broadband VIs to augment NDVI. We have developed multiple PIC Spectral Imaging Indices (PIC-SIIs) to provide you the most precise and useful information available. PIC-SIIs lend insight into field health by utilizing spectral information in such a way that soil effects can be minimized, subtle changes within a crop can be detected and yield estimates can be fine-tuned. Here are just a few examples of other broadband VIs which are useful for precision agriculture applications.
Atmospherically Resistant VI (ARVI):
This index is an enhancement to NDVI that is relatively resistant to atmospheric factors (for example, aerosol). It uses blue reflectance to correct red reflectance for atmospheric scattering. It is most useful in coastal regions which often have high levels of atmospheric aerosol.
Enhanced Vegetation Index (EVI):
This index was developed to improve NDVI by using the blue reflectance region to correct for soil background signals and reduce atmospheric influences, including aerosol scattering. It is most useful in regions where NDVI often saturates.
This index is similar to NDVI except that it substitutes the green spectrum (540 to 570 nm) for the red spectrum. It is more sensitive to chlorophyll concentration than NDVI.
NDVI and other broadband indices utilize only a fraction of the information available in the original spectral reflectance data captured by the advanced PIC HSI sensor. It was not until the mid-1990s that a new generation of algorithms were developed to take advantage of the enhanced performance and characteristics of modern sensors such as the PIC VNIR HSI sensor. In this section we will review these improved VIs by describing general categories and providing a short description of some of the most used indices. We will describe the more targeted assessments of vegetation vigor/stress that our HSI sensor enables.
Narrowband VIs require high spectral resolution provided by hyperspectral imaging spectrometers. They provide a measure of the overall amount and quality of photosynthetic material in vegetation.
Most narrowband VIs are derived from high spectral resolution measurements in the red and near-infrared (NIR). The high spectral resolution achieved with the PIC sensor (~2 nm) is used to precisely sample the red edge and NIR portions of the reflectance curve (Fig. 2). Because our high resolution NIR measurements penetrate more deeply into the canopy than visible-only measurements we can determine the total amount of green material throughout the dense canopy.
By making narrowband measurements near the red edge, PIC-SIIs are sensitive to smaller changes in vegetation health than broadband greenness VIs (e.g. NDVI). This is particularly relevant in conditions of dense vegetation (such as vineyard canopies) where simple broadband VI values can saturate and lead to false conclusions.
Here is a sample set of narrowband VIs that PIC provides now.
Red Edge Position (REP) – PIC SII 3.0
is a narrowband measurement sensitive to changes in chlorophyll concentration. Increased chlorophyll concentration moves the red edge to longer wavelengths. The high spectral resolution of the PIC VNIR HSI sensor allows us to monitor the red edge position very accurately. This index calculates the wavelength of the maximum derivative (highest rate of change) of reflectance in the red edge region. REP is useful for crop monitoring by giving you advance notice of canopy stress caused by climate and other factors.
Vogelmann Red Edge Index (VOGRE) – PIC SII 5.0
is a narrowband reflectance measurement sensitive to the combined effects of chlorophyll concentration, canopy leaf area and water content. PIC SII 5.0 is used for vegetation phenology (growth) studies and crop productivity estimations.
Vegetation reflectance indices can also be categorized by utility. There are classes of newly developed indices that our high resolution VNIR and future SWIR sensor will make available to improve your efficiency.
Light Use Efficiency Indices
These indices quantify your crop's ability to use incident light for photosynthesis. Only a small range of the electromagnetic spectrum is utilized by plants during photosynthesis. This range happens to fall within the visible portion of the spectrum, from 400nm to 700nm. A plant's ability to efficiently absorb energy within this range can be a good predictor of growth rate and biomass production. PIC is working to provide three vegetation indices to measure light use efficiency:
Photochemical Reflectance Index (PRI) – PIC SII 6.0
is particular useful for measuring vegetation health prior to senescence.
Structure Insensitive Pigment Index (SIPI):
maximizes sensitivity to the ratio of bulk carotenoids to chlorophyll while minimizing the impact of the variable canopy structure.
Red Green Ration Index (RGRI):
useful for making foliage development estimations, indicating leaf production and stress, or even indicating flowering in some canopies. The ratio measures the relative expression of leaf redness caused by anthocyanin to that of chlorophyll.
Dry/Senescent Carbon Indices
These season indices exploit characteristics found in vegetation components during senescence including changes to lignin (used to make woody stems) and cellulose (used for cellular tissue structure). The concentration of these materials increase when vegetation is about to undergo senescence.
Plant Senescence Reflectance Index (PSRI)
is sensitive to the ratio of bulk carotenoids to chlorophyll. An increase in PSRI indicates increased canopy stress (carotenoid pigment), the onset of canopy senescence and plant fruit ripening. PSRI is used for vegetation health monitoring, stress detection, crop production and yield analysis.
Leaf Pigment Indices
These are indices designed to provide a measure of stress-related pigments present in vegetation. Stress-related pigments, including carotenoids (yellow pigments) and anthocyanins (pink, purple and red pigments), tend to be present in higher concentrations when vegetation is in a weakened state. Carotenoids function in light absorption processes in plants as well as in protecting plants from the harmful effects of high light conditions. Anthocyanins are water-soluble pigments abundant in newly-formed leaves and leaves undergoing senescence. The leaf pigment vegetation indices do not measure chlorophyll. Applications for leaf pigment vegetation indices include analyses of canopy stress and ecosystem studies.
Carotenoid Reflectance Index (CRI)
is sensitive to carotenoid pigments in plant foliage. Estimating carotenoid content is more difficult than estimating chlorophyll because chlorophyll and carotenoid absorption peaks overlap and the concentration of carotenoid is much lower in most leaves.
Canopy Water Content Indices
These indices are used to measure the amount of water contained in the foliage canopy. Higher water content usually indicates healthier vegetation likely to grow faster. Canopy water content vegetation indices use reflectance measurements in the near-infrared and shortwave infrared regions to take advantage of known absorption features of water and the penetration depth of light in the near-infrared region to make integrated measurements of total column water content. Canopy water content indices are able to penetrate into thick canopies.
Water Band Index (WBI)
– PIC SII 8.0 is sensitive to changes in canopy water content. The PIC sensor measures the strength of absorption (reduction in reflectance) at 970nm. PIC SII 8.0 is useful for canopy stress analysis, productivity predictions and studies of ecosystem physiology.
Moisture Stress Index (MSI):
is a narrowband index that requires information in the 1,599 nm (SWIR) range which is not provided by most sensors used in agricultural applications. As water content increases spectral absorption at 1,599 nm increases. PIC will be adding a SWIR HSI sensor in the near future.
We continue to work with University and government agricultural scientists to develop targeted indices addressing specific issues such as mildew (MI – PIC SII 9.0), grape leafroll and redblotch disease (GLD/GRD – PIC SII 10.0), Botrytis Cinerea (BCI – PIC SII 16.0 and maladies. We are also developing indices that will assess fertilizer coverage (FCI PIC-SII 17.0) and a crop ripeness Fruit Sugar Index (FSI PIC-SII 11.0). PIC will remain at the forefront of remote sensing for precision agriculture by continuing our close collaboration with University and government agricultural scientists.
Vegetation Indices will never measure exact concentrations or abundance of any agriculture component or parameter. PIC VIs (aka PIC SIIs) will enable you to monitor your crops more efficiently and make strategic decisions on a block by block or vine by vine basis.
Each block, varietal or field condition has an impact on the reflectance measurements at the heart of VIs. Some VIs will provide better information than others in any given area. PIC will assess which indices are most appropriate for the blocks, varietals and field conditions of your vineyard by comparing results of different VIs and correlating these to measured on-site conditions.
By offering multiple VIs and working with University and government scientists PIC will provide a product tailored to your unique needs and conditions. What PIC offers right now is a vast improvement over current NDVI maps. Our data and vine-level imagery will improve your strategic decisions, increase your crop yield and reduce your operating costs.