Vegetation Red Edge Index Based on Hyperspectral Data: Unlocking the "Spectral Code" of Crop Health

In the field of agricultural remote sensing, hyperspectral technology reveals the health status of vegetation by capturing fine spectral information of crop reflectance. The red edge (680-750 nm) is a rapid transition region in the vegetation spectrum from low red light reflectance to high near-infrared reflectance, and is extremely sensitive to chlorophyll content, stress conditions, and growth stages. This article will introduce six red edge indices (REP, NDVIre, RedEdgeSlope, RedEdgeArea, mSRre, and RedEdgeCurvature) based on hyperspectral data spectral curves, including their calculation methods, application scenarios, and comparisons with traditional vegetation indices (such as NDVI and SAVI), revealing the unique advantages of hyperspectral indices.

I. Calculation Method and Application of Red Edge Index

The red-edge index utilizes fine spectral data from hyperspectral satellites to uncover subtle changes in the red-edge region. Below is a detailed introduction to six red-edge indices:

1. REP (Red Edge Position)

Calculation method:

REP is the wavelength corresponding to the maximum value of the first derivative (slope) of the spectrum in the red edge region (680-750 nm), reflecting the position of the spectral inflection point.

step:

1) Calculate the first derivative of the spectral reflectance.

2) Find the wavelength corresponding to the maximum derivative in the red-edge region.

3) Use linear interpolation to refine wavelength accuracy.

Application scenarios:

Crop health monitoring: REP is usually at 710-720 nm (healthy vegetation). If it shifts to shorter wavelengths (

Species classification: Different crops (such as wheat and corn) have different REP positions, which can be used to distinguish crop types.

Yield forecasting: REP is related to photosynthetic efficiency and can be used to indirectly estimate crop yield.

Example: In a wheat field, a shift in REP from 715 nm to 700 nm may indicate drought stress, requiring increased irrigation.

2. NDVIre (Red-edged Normalized Difference Vegetation Index)

Calculation method:

The normalized index is calculated using reflectance at 705 nm (red edge starting point) and 750 nm (near infrared): NDVIre = (R750 - R705) / (R750 + R705)

Application scenarios:

Vegetation cover assessment: A high NDVIre value (close to 1) indicates dense vegetation, while a low value (

Early stress detection: Sensitive to water or nutrient stress, suitable for early warning.

Example: A drop in NDVIre from 0.8 to 0.4 may indicate insufficient nitrogen fertilizer, requiring fertilization.

3. RedEdgeSlope

Calculation method:

Calculate the maximum value of the first derivative of the spectrum in the red edge region (680-750 nm), which represents the transition speed of the spectrum from red light to near-infrared.

Application scenarios:

Stress diagnosis: A decrease in slope indicates reduced chlorophyll or damage to leaf structure (such as by pests or diseases).

Growth stage monitoring: The slope changes with the crop growth stage, which is suitable for phased management (such as seedling stage and flowering stage).

Example: A significant decrease in the red edge slope during the wheat flowering period may indicate disease and require pest and disease control.

4. RedEdgeArea

Calculation method:

The integral area of ​​the spectral reflectance in the red-edge region (680-750 nm) can be calculated using methods such as the trapezoidal integral.

Application scenarios:

Biomass estimation: Area is positively correlated with vegetation cover and biomass.

Stress monitoring: A reduction in area may reflect drought or disease.

Ecological research: assessing vegetation photosynthetic capacity and carbon sequestration potential.

Example: Areas with reduced red border area may be drought-affected farmland that require priority irrigation.

5. mSRre (Improved Red Edge Simplified Ratio)

Calculation method:

Combining 445 nm (blue light), 705 nm (red edge), and 750 nm (near-infrared): mSRre = (R750 - R445) / (R705 - R445)

Application scenarios:

Chlorophyll estimation: Sensitive to chlorophyll absorption, suitable for monitoring nutritional status.

Stress detection: A decrease in mSRre may indicate nitrogen deficiency or disease.

Crop classification: The mSRre values ​​of different crops vary significantly.

6. Red Edge Curvature

Calculation method:

Calculate the average absolute value of the second derivative in the red-edge region. The second derivative reflects the shape change of the spectral curve.

Application scenarios:

Early stress detection: Curvature is sensitive to subtle physiological changes, such as early disease.

Vegetation health assessment: related to chlorophyll and leaf structure.

Dynamic monitoring: Tracking changes in crops during the growing season.

Example: Abnormal fluctuations in curvature may indicate an early fungal infection, requiring ground examination.

II. Comparison with traditional vegetation indices

Traditional vegetation indices (such as NDVI and SAVI) are widely used in multispectral remote sensing, but red-edge indices based on hyperspectral data spectral curves have significant advantages. Vegetation indices calculated from wide-band multispectral data often overlook subtle changes in the red-edge region, while narrow-band hyperspectral data can accurately capture the spectral characteristics of this region. For example, REP locates inflection points using the first derivative, and NDVIre, using narrow bands at 705 nm and 750 nm, offers higher sensitivity. It is also more sensitive to subtle changes in chlorophyll content and early stresses (such as drought and disease).

Meanwhile, red-edge indices based on spectral curves also have stronger stress detection capabilities. Indices such as NDVI tend to saturate when vegetation is dense and are not sensitive to early stress. While SAVI improves soil impact, it still relies on a wide wavelength band. NDVIre and mSRre, on the other hand, directly use the red-edge band and are sensitive to chlorophyll changes. REP and RedEdgeCurvature can detect changes in spectral shape, making them suitable for early stress warning and enabling earlier detection of drought or nutrient deficiency (such as REP shifting to shorter wavelengths), thus buying time for agricultural management. By combining various types of vegetation indices, crop status can be assessed more comprehensively, improving the accuracy of stress classification.

The following is the vegetation red edge index calculated using the XIGUNAG-002 hyperspectral satellite. The data was acquired on April 5, 2025, in Yuncheng City, Shanxi Province. The satellite has a swath width of 80 km, a wavelength range of 430-850 nm, 150 bands, and a ground resolution of 40 meters.

Red-bordered area (April 5, 2025, Yuncheng City, Shanxi Province)

Red-edged normalized vegetation index (April 5, 2025, Yuncheng City, Shanxi Province)

Improved red-edge simplified ratio (April 5, 2025, Yuncheng City, Shanxi Province)

The hyperspectral red-edge index provides a powerful tool for crop health monitoring by mining fine spectral information within the red-edge range. Compared to traditional vegetation indices, the red-edge index has higher spectral resolution, stronger stress detection capabilities, and richer physiological information, making it suitable for early stress warning and precision management in precision agriculture. In the future, with the widespread adoption of hyperspectral satellites, the red-edge index will shine in smart agriculture, providing farmers with more accurate decision support.