Hyperspectral Image & Index Calculator for Water Contamination

An expert tool to identify the correct spectral bands and indices for monitoring water quality parameters like algae, sediments, and dissolved organic matter.

Water Quality Parameter Selector

What is Hyperspectral Water Contamination Analysis?

Hyperspectral water contamination analysis is a remote sensing technique used to assess water quality over large areas. Unlike standard cameras that see light in three bands (Red, Green, Blue), hyperspectral sensors capture data across hundreds of narrow spectral bands, extending from the visible into the near-infrared and short-wave infrared spectrums. This creates a detailed “spectral signature” for every pixel in an image. Different substances in water, such as algae (chlorophyll), sediment, and dissolved organic matter, absorb and reflect light differently. By analyzing these unique spectral signatures, scientists can identify the presence and estimate the concentration of these key water quality indicators without direct water sampling. This method is invaluable for monitoring large lakes, reservoirs, and coastal areas, providing insights into ecological health, pollution events, and the spread of harmful algal blooms.

Hyperspectral Index Formulas and Explanation

To quantify water contaminants, researchers use spectral indices. These are simple mathematical formulas that combine the reflectance values of two or more spectral bands to enhance the signal of a specific substance while minimizing interference from others. For instance, an index for chlorophyll will use bands where chlorophyll reflects strongly and where it absorbs strongly.

Common Water Quality Indices and Their Variables

Variable	Meaning	Unit	Typical Spectral Range
R(Blue)	Reflectance in the Blue band	Unitless (reflectance ratio)	450 – 510 nm
R(Green)	Reflectance in the Green band	Unitless (reflectance ratio)	500 – 590 nm
R(Red)	Reflectance in the Red band	Unitless (reflectance ratio)	620 – 680 nm
R(Red Edge)	Reflectance in the Red Edge transition zone	Unitless (reflectance ratio)	690 – 740 nm
R(NIR)	Reflectance in the Near-Infrared band	Unitless (reflectance ratio)	760 – 900 nm

Practical Examples

Example 1: Detecting an Algal Bloom

A reservoir manager suspects an algal bloom is forming. They use the calculator and select Chlorophyll-a. The tool recommends the Normalized Difference Chlorophyll Index (NDCI). Using recent hyperspectral imagery (e.g., from Sentinel-2), they apply the formula: NDCI = (R(Red Edge) - R(Red)) / (R(Red Edge) + R(Red)). Areas with high positive values indicate dense concentrations of chlorophyll, confirming the location and extent of the bloom and allowing for targeted management action. This is a common application discussed in studies on chlorophyll-a concentration estimation.

Example 2: Monitoring River Sediment After a Storm

After a heavy rainfall event, a government agency needs to assess how much sediment has been washed into a major river system. They select Total Suspended Solids (TSS) in the calculator. A common approach is to use the reflectance in the Red or Near-Infrared (NIR) band, as sediment increases scattering in these regions. A simple algorithm might be a single band model using R(NIR). High reflectance values in the river channel on satellite imagery correlate strongly with high TSS concentrations, helping to identify areas of significant erosion and deposition. Many studies focus on retrieving total suspended solids from imagery.

How to Use This Hyperspectral Index Calculator

1. Select Parameter: Choose the water quality parameter you are interested in from the dropdown menu (e.g., Chlorophyll-a, TSS).
2. Calculate: Click the “Calculate Recommended Indices” button.
3. Review Results: The tool will display the most common and effective spectral index for your selection, along with its mathematical formula.
4. Identify Bands: Note the specific spectral bands (e.g., Red Edge, NIR) required for the calculation. The approximate wavelengths are provided to help you match them with your sensor’s specifications.
5. Check Satellites: A list of common satellite missions whose sensors (like Sentinel-2 MSI) provide the necessary bands is shown. You can find more information about hyperspectral imaging satellites online.
6. Interpret Chart: The spectral chart provides a visual guide to the reflectance pattern of your chosen substance, helping you understand why certain bands are used in the formula.

Key Factors That Affect Water Quality Measurement

• Atmospheric Conditions: Haze, clouds, and aerosols scatter light and must be corrected for. This is why Bottom-Of-Atmosphere (BOA) reflectance is preferred over Top-Of-Atmosphere (TOA).
• Sun Angle & Viewing Geometry: The position of the sun and the sensor can cause glare (sun glint) or shadows, affecting reflectance values.
• Water Depth and Bottom Reflectance: In very clear, shallow water, the reflection from the lake or riverbed can interfere with the water column signal.
• Co-existing Substances: Water is a complex mixture. High levels of sediment can interfere with chlorophyll measurements, and high CDOM can darken the water, affecting all measurements. Advanced algorithms are needed to untangle these signals.
• Sensor Resolution: Both the spatial (pixel size) and spectral (number and width of bands) resolution of the sensor determine the level of detail that can be captured. Hyperspectral sensors are superior to multispectral ones for this reason.
• In-situ Data for Calibration: Remote sensing algorithms provide the best results when they are calibrated and validated with physical water samples collected at the same time as the image acquisition (in-situ data).

Frequently Asked Questions (FAQ)

What is the difference between hyperspectral and multispectral imaging?

Multispectral imaging (like that from Landsat or Sentinel-2) captures data in a few, wide spectral bands (e.g., 5-15 bands). Hyperspectral imaging captures data in hundreds of very narrow, contiguous bands, providing a much more detailed spectral signature. This detail allows for the identification of specific materials that might be indistinguishable with multispectral data. For more details, see how remote sensing evaluates water quality.

Can this calculator give me the exact contamination level in mg/L?

No. This tool identifies *how* to perform the calculation by recommending the right index and spectral bands. To convert an index value (like NDCI) into an absolute concentration (e.g., mg/L of chlorophyll-a), you need to develop an empirical model by correlating index values from the image with measurements from physical water samples taken in the field.

What does CDOM stand for?

CDOM stands for Colored Dissolved Organic Matter. It’s a complex mix of organic substances, primarily from decaying plant and animal matter, that leach into water and give it a yellow or brown “tea-stained” color. It strongly absorbs blue light.

What are Total Suspended Solids (TSS)?

TSS refers to all solid particles that are suspended (not dissolved) in the water column. This includes inorganic materials like silt and clay, and organic materials like algae and detritus. High TSS makes the water turbid or cloudy.

Why is the Red Edge band so important for vegetation and algae?

The Red Edge is a region in the spectrum (~690-740 nm) where the reflectance of vegetation and algae increases very sharply. Chlorophyll absorbs strongly in the red region (~670 nm) and the internal cell structure reflects very strongly in the near-infrared (~760 nm+). The steep slope between these two points, the Red Edge, is highly sensitive to changes in chlorophyll concentration. This is crucial for spectral signatures analysis.

Which satellite is best for water quality monitoring?

There is no single “best” satellite. Sentinel-2 from the Copernicus program is very popular because it is free, has a high revisit time (5 days), and includes key bands like the Red Edge. However, true hyperspectral satellites like PRISMA, EnMAP, or commercial options from companies like Pixxel offer far more spectral detail for advanced research. Research on UAV-borne hyperspectral imagery also shows promise for very high-resolution monitoring.

Do I need to perform an atmospheric correction?

Yes, for accurate quantitative analysis, an atmospheric correction is critical. It converts the Top-of-Atmosphere (TOA) reflectance measured by the satellite into Bottom-of-Atmosphere (BOA) or surface reflectance by removing the effects of atmospheric scattering and absorption. Some indices are designed for TOA data for rapid assessment, but BOA is the standard for scientific work.

What does the spectral signature chart show?

The chart shows a representative “fingerprint” of how a substance interacts with light. For example, chlorophyll has characteristic absorption peaks in the blue (~450 nm) and red (~670 nm) parts of the spectrum, and higher reflectance in the green and near-infrared. This pattern is what allows algorithms to detect it.