Assessment of Lake Algae Blossoms and Chlorophyll Concentration

Monitoring the concentration of algal blooms and chlorophyll-a (Chl-a) in lakes, forEutrophication management, protection of drinking water sources, and assessment of ecosystem healthcrucial.When blue-green algae reproduce in large numbers, harmful toxins are produced, posing a threat to drinking water supply and aquatic life.Therefore, as forChlorophyll a as a substitute indicator for algal biomassAccurate and spatially continuous monitoring is crucial for environmental management and water quality modeling.The traditional on-site sampling method can provide accurate point measurement data, but the spatial coverage is sparse and labor-intensive.By contrast,Remote sensing technology can achieve large-scale, repeatable, and scalable assessments.Among various remote sensing technologies,Hyperspectral imaging can provide the most detailed spectral information, even in inland waters with complex optical properties, it can reliably distinguish algal pigments and estimate chlorophyll concentration.

Hyperspectral imaging (HSI) can be used to measure water quality parametersCross space and timeDetailed spectral characterization.

thistextfocus onGround hyperspectral imaging system，This system can provide high-resolution, flexible, and cost-effective monitoring of lakes and reservoirs.by means ofFixed or mobile platforms near the water surfaceOperating on (such as docks, monitoring towers, or ships), ground hyperspectral imaging is availableFilled the gap between in-situ measurement and aerial/satellite observation.

Principles of Hyperspectral Remote Sensing

Hyperspectral sensors can obtain400-1000 nanometersIn the visible to near-infrared wavelength rangeHundreds of consecutive narrow spectral bands (typically with a bandwidth of 2-10 nanometers)ofReflectance dataThis fine spectral resolution canAccurately identify subtle absorption and scattering characteristics related to pigments, suspended solids, and dissolved organic substances.

The upward off water reflectance spectrum of lake water is a composite signal that is influenced by various factors, including:

l The absorption of chlorophyll a and auxiliary pigments such as phycocyanin and carotenoids

l Backscatter effect of suspended sediment and phytoplankton

l Absorption of Colored Dissolved Organic Matter (CDOM)

l Fluorescence emission of chlorophyll near 681 nanometers

The ground system can achieve jiHigh Signal to Noise Ratio (SNR)Analyze these fine scale features to make them an ideal choice for calibration and validation research.

Advantages of Ground Hyperspectral Imaging

l Controllable observation geometry

Ground systems (installed on docks, ships, or tripods) can be used forObservation angle and lighting anglePerform precise control,maximumlimitReduce specular reflection and neighborhood effects.

l High spatiotemporal resolution

Can be capturedCentimeter to meter scaleLocalized algal bloom patches or gradient changes.

Repeated collection (minute to hour level)Capable of conducting time series analysis on the evolution of algal blooms.

l Align directly with in-situ measurements

Can be easily combined with water sample collection (chlorophyll a, phycocyanin, total suspended solids (TSM), and colored dissolved organic matter (CDOM)).

Algorithm validation that facilitates satellite or drone applications.

l Cost effectiveness and accessibility

Avoid issues related to aircraft costs, flight logistics, and airspace permits.

Suitable for continuous or semi permanent monitoring stations.

l Flexibility of Spectral Configuration

Portable spectrometer or hyperspectral cameraIt can be adjusted to the visible light or visible near infrared (VNIR) wavelength range according to the target pigment.

Configuration of Ground Hyperspectral Imaging System

² Typical components

l Hyperspectral camera (push scan type)

l Stable installation platform (tripod, gimbal or universal joint)

l Calibration accessories (Spectralon board, reference lamp for radiometric calibration)

l Data acquisition computer with GPS/time marker

l Optional downlink irradiance radiation sensor (for calculating reflectance)

² Installation Options

l Fixed station: installed on docks, monitoring towers, or bridges for repeated measurements

l Mobile platform: mounted on ships or floating rafts to scan the cross-section of lakes

l Scanning settings: Horizontal scanning of certain areas of the lake to generate hyperspectral mosaic images

Key spectral features for algae and chlorophyll detection

The advantages of hyperspectral imaging in inland waters

l Enhance pigment identification ability

Hyperspectral data is analyzableNarrow absorption characteristics(For example, the 620 nanometer absorption peak of phycocyanin)，Thus achieving the distinction between blue-green algae and green algae.

l Improve the accuracy of chlorophyll quantification

Narrow bandindexCan capture subtle red edge offsetsIt can achieve chlorophyll estimation in both nutrient poor and nutrient rich water bodies.

l Algorithm design flexibility

Users can adjust custom band combinations or apply semi analytical models,Unrestricted to fixed multispectral bands.

l Cross sensor transferability and machine learning training

Hyperspectral datasets support the development of machine learning models such as random forest (RF), edge gradient boosting (XGB), and convolutional neural networks (CNN)Can generalize across different lakes and seasons.

l prospective

The new satellite missions (PRISMA, DESIS, EnMAP, CHIME) and aviation sensors ensureContinuity and global coverage of data.

On site implementation example

Actual deployment may include the following steps:

l A ClydeHSI VNIR-S camera (with a wavelength range of 400-1000 nanometers and a spectral resolution of 5 nanometers)pretendAt the pier overlooking the lake

l Periodic imaging is performed every 30 minutes during the daytime period

l Simultaneous collection of water samples for determination of chlorophyll a (Chl-a), phycocyanin (PC), and total suspended solids (TSM)

l Calibrate using a Spectralon board with a reflectivity of 99%

l Process data to draw a distribution map of chlorophyll-a in the nearshore area, with a resolution of approximately 10 centimeters.

This type of system can detect early occurrence of algal blooms, track daily changes in pigments, and provide real ground data for satellite algorithm validation.

Process of using hyperspectral data

l data collection

Obtain hyperspectral imagery (such as PRISMA, DESIS, EnMAP satellite or aerial imagery)

Ensure that the collection time is synchronized with the field sampling time used for calibration

l preprocessing

Perform radiometric calibration and atmospheric correction to derive the water free reflectance (ρ w or Rrs)

Performing glare and proximity effect correction (crucial for small lakes)

l spectral analysis

Extracting spectra of water pixels using region of interest (ROI) masks or shapefiles

Calculate spectral indices (such as NDCI, MCI, phycocyanin (PC) ratio)

Optional steps: Perform derivative analysis or continuum removal to enhance spectral features

l Algorithm application

Apply optimized empirical models or machine learning regression models trained on local field data

Generate concentration distribution maps of chlorophyll and/or phycocyanin

l Verification and Calibration

Compare the concentration obtained from satellite inversion with in-situ chlorophyll-a data

Evaluate accuracy using root mean square error (RMSE), deviation, and coefficient of determination (R ²)

l output

Generate a geographic reference map of chlorophyll-a and phycocyanin (PC) concentrationsIdentify algal bloom areas and their temporal changes to provide support for management response

Example: Application of Hyperspectral Data in Chlorophyll Estimation of Lakes

1. Extract reflectance spectra from hyperspectral images of lake areas

2. Calculate the Normalized Chlorophyll Index (NDCI) or Three Band Red Edge Index for each pixel

3. Using regression coefficients derived from field data, convert the index value into chlorophyll-a concentration

4. Visualize spatial distribution to identify areas of algal bloom intensity

This process can achieve near real-time monitoring of algal bloom dynamics and facilitate integration with hydrodynamic or water quality models.

Optional: UAV based hyperspectral imaging system

if neededCovering a larger lake areaThe unmanned aerial vehicle (UAV) hyperspectral system can provide aFlexible intermediate solutions

Modern lightweight push scan or snapshot cameras mounted on drones, such as Headwall Nano Hyperspec

Cubert UHD can:

l obtainCentimeter level resolutionSpectral data

l InWithin a few minutesCovering the entire surface of the lake

l Support and foundation systemSame calibration and processing procedures

However, drone operations require obtaining airspace permits, ensuring stable lighting, and precise radiometric calibration to ensure resultsQuantification.

Ground based hyperspectral imaging provides a powerful, flexible, and cost-effective method for monitoring lake water quality. Hyperspectral imaging provides a way to monitor the dynamics of algal blooms and chlorophyll in lakesRich and scalable quantitative and spectral informationThe solution. Its narrowband dataCapable of capturing key pigment absorption and scattering characteristics missed by multispectral systemsThus, precise detection of algal biomass and blue-green algae activity can be achieved.

The main advantages include:

l High temporal resolution and accuracy of local areas

l Can directly verify biological optical models and machine learning models

l Suitable for long-term deployment or automated deployment

l Potential for integration with unmanned aerial vehicles (UAVs) and satellite systems

By focusing on hyperspectral observations of the foundation, researchers and managers can establishContinuous and quantitative pigment monitoringTesting frameworkTo provide support for harmful algal bloom warning systems and reliable calibration basis for a wider range of remote sensing networks.