Remote Sensing and Vegetation Indices

Illumination Sources and Sensors

  • Two sources of illumination:
    • Sun.
    • Satellites, airplanes, or drones with their own light source.
  • Active Sensors: Sensors on vehicles with their own light source.
  • Passive Sensors: Sensors that rely on the sun as the primary source of illumination.
  • Radiation Interaction:
    • Solar radiation hits the Earth's surface and interacts with the atmosphere and the surface.
    • Interactions with the surface: absorption, reflection, and transmission.
  • Atmospheric Attenuation:
    • Radiation undergoes two-way atmospheric attenuation (coming and going back).
    • These effects must be removed during preprocessing.
  • Radiometric Resolution:
    • Satellites convert reflected radiation into digital numbers (DN) through quantization.
    • Remote sensing analysts analyze these DN numbers to produce useful information.

Reflectance

  • Reflectance is the ratio of reflected electromagnetic radiation to incident radiation in passive sensors or optical remote sensing.
  • Reflectance=Reflected Electromagnetic RadiationIncident RadiationReflectance = \frac{Reflected \, Electromagnetic \, Radiation}{Incident \, Radiation}
  • There is a dialectic relationship between reflected and incident radiation.

Spectral Indices (Image Indices)

  • Spectral indices are images computed from multiband images.
  • Landsat 8: Each spectral band is one image output.
  • Spectral indices emphasize specific phenomena and mitigate other degrading effects.
  • Definition: Combinations of spectral reflectance from two or more wavelengths that indicate the relative abundance of a feature of interest.
  • Example: NDVI emphasizes vegetation, displaying it in brighter colors in a grayscale output.

Importance of Studying Vegetation Indices

  • Vegetation is an important land use and land cover.
  • Satellite data shows global vegetation distribution in different months (e.g., July and August).
  • Vegetation indices help in understanding the spatial and temporal distribution of vegetation.
  • Example: Normalized Difference Vegetation Index (NDVI) collected by MODIS on NASA's STERRA satellite.

Spectral Signatures

  • Every material has a distinctive spectral signature, which is the reflectance of a surface in different wavelengths.
  • Example:
    • Green line: Vegetation.
    • Red line: Soil.
    • Blue line: Water.
  • Water absorbs radiation more effectively, resulting in less reflectance.
  • Experienced remote sensing analysts can identify materials by examining spectral curves.
  • Spectral signature: Plots of wavelength against reflectance.

Wavelength Regions

  • X-axis: Wavelength in micrometers.
  • Visible Region: 0.4 to 0.75 micrometers.
    • Blue: 0.45 to 0.495 micrometers.
    • Green: 0.495 to 0.625 micrometers.
    • Red: 0.62 to 0.75 micrometers.
  • Near Infrared (NIR): 0.75 to 1.2 micrometers.
  • Mid Infrared (MIR): 1.3 to 2.6 micrometers.
  • Optical Remote Sensing focuses on these regions.

Vegetation Reflectance

  • Typical Spectral Signature of Vegetation (Green Line):
    • Absorption in blue and red bands (visible region).
    • Less absorption in the green band (hence, green color).
    • High reflectance in the near-infrared (NIR) region.
  • High Absorption Areas:
    • Around 1.5 and 1.95 micrometers due to atmospheric water absorption bands.
  • Vegetation Behavior in Wavelengths:
    • Visible: Generally absorbed.
    • Near Infrared: Mostly reflected.
    • Middle Infrared: More absorption.

Physiological Causes of Spectral Signature

  • Visible Region (0.4 to 0.75 micrometers):
    • Absorption due to chlorophyll, essential for photosynthesis.
    • Plants absorb sunlight and combine carbon dioxide and water to produce carbohydrates and release oxygen.
    • Blue and red parts of the visible spectrum have more absorption.
  • Near Infrared (NIR):
    • High reflectance due to scattering of light.
    • Empty spaces and pores inside leaves facilitate gas exchange (carbon dioxide and oxygen).
    • These structures cause scattering of NIR radiation.
  • Middle Infrared:
    • More absorption due to leaf water content.
    • Dry or burnt leaves reflect more in this region as they lose water content.

Summary of Leaf Reflectance

  • Leaf pigment absorption in the visible channel (0.4 to 0.75 micrometers).
  • Internal scattering in the near-infrared (0.7 to 1.3 micrometers) due to spongy mesophyll.
  • Leaf absorption in the middle infrared regions.

Spectral Signatures of Different Surfaces

  • Different surfaces have distinctive reflectance curves.
  • Examples:
    • Pinewood spectral signature.
    • Grassland.
    • Red sand pit.
    • Silty water (more reflectance than pure water).
  • Spectral signatures are important for identifying different surfaces.
  • Understanding spectral indices is necessary to distinguish between different Earth surfaces in space and time.

Temporal Domain and Spectral Indices

  • Vegetation indices, such as Normalized Difference Vegetation Index (NDVI), can be studied over time.
  • Example: NDVI curves over one year using SPOT vegetation data.
  • NDVI values increase in winter (Northern Hemisphere) for winter crops like rain-fed wheat and sunflower, peaking in March and April.
  • NDVI is good for mapping crop areas but may not be as effective for estimating crop biomass or yield due to saturation.

Soil Moisture Content and Reflectance

  • Soil with different moisture contents shows varying spectral curves.
  • Example: Syltrum soil.
    • Less moisture content: Higher reflectance.
    • More moisture content: More absorption, lower reflectance.
  • Reflectance increases as moisture content decreases.

Spectral Indices and Mathematical Combinations

  • Spectral indices are mathematical combinations or transformations of bands.
  • They accentuate spectral properties and features.
  • Example: Transforming spectral bands to highlight vegetation or water.
  • Many spectral indices exist, including vegetation indices, geology indices, and landscape indices.
  • Clay mineral ratio index: Ratio between shortwave infrared bands (SWIR1/SWIR2 for Landsat 8).
  • Burn ratio index: A landscape index used in practical six.

Vegetation Indices Characteristics

  • Vegetation indices are dimensionless radiometric measurements that indicate the relative abundance and activity of green vegetation.
  • Characteristics of a good vegetation index:
    • Maximize sensitivity to plant biophysical parameters.
    • Normalize external effects such as sun angle, viewing angle, and shadows.

Characteristics of Good Vegetation Indicator

  • Wavelength Specificity: Reflectance in a particular wavelength that maximizes sensitivity.
  • Normalization: Normalize illumination characteristics of the remote sensing system, atmospheric impacts, and internal effects like canopy background variations.
  • Example: Soil Adjusted Vegetation Index (SAVI) is better than NDVI for areas with sparse plants because it normalizes the impact of soil background.
  • Coupling With Biophysical Parameters: Relate to biomass or leaf area index.
  • Rate of Reflectance: Dependent on wavelength, type of surface, illumination, and atmosphere.

Vegetation vs Soil

  • Vegetation reflects differently from soil.
  • Red edge: Peak in reflectance between red and near-infrared bands can discriminate vegetation.
  • Factors: Chlorophyll absorption, plant structure scattering, and liquid water absorption.
  • Multispectral Bands: Zoom into different surfaces using individual bands to understand vegetation indices.

Vegetation Indices: Band Ratios

  • Vegetation indices are combinations of spectral reflectance from two or more wavelengths.
  • Band ratios are quotients between measurements of reflectance in separate portions of the spectrum.
  • Involve at least two spectral bands (e.g., near infrared divided by red).
  • Understanding the outcome of dividing one band by another helps in discriminating land cover types.

Vegetation Indices - Examples

  • Difference Vegetation Index (DVI): Simple difference of near-infrared and red.
    • Sensitive to biomass.
    • Distinguishes between soil and vegetation.
    • Does not account for atmospheric or shadow differences.
  • Simple Ratios (NIR/Red):
    • Low values for soil, water, or ice.
    • High values for healthy vegetation.
    • Potential issue: Red can be zero, needs smart handling.
  • Considerations before using ratios or differences:
    • Type of surface, context of the image, expected outcome, and purpose (mapping vs. biophysical analysis).

Normalized difference vegetation index (NDVI)

  • Most heavily used index for vegetation studies.
  • Formula: NDVI=NIR−RedNIR+RedNDVI = \frac{NIR - Red}{NIR + Red}
  • Ranges from -1 to 1.
    • -1 to 0: Other surfaces like water, clouds, snow (good absorbers).
    • More than 0: Vegetation.
  • Ideal Framework:
    • Download image, preprocess, rescale reflectance data, and calculate vegetation indices.
  • Strengths and Limitations:
    • Strengths: Monitors seasonal changes in vegetation growth, reduces noise in temporal images, part of operational decision support systems for food security.
    • Limitations: Saturates when biomass is high, sensitive to canopy background, and issues with pest attacks or diseases.

Shortwave infrared

  • Uses near-infrared and shortwave infrared.
  • Relates to plant water content.
  • Values range from -1 to 0.
    • Less than 0: No water content.
    • More values: More water content.
  • Modified Normalized Difference Vegetation Index: Uses green and near-infrared.
    • More common in open water such as floods.
  • Interpretation:
    • Less than 0.3: No open water.
    • Greater than 0.3: Open water (standard value, depends on images).