TL;DR
This paper introduces an Automatic Stationary Water Color Parameters Observation System (AFWCPOS) designed for continuous, high-frequency monitoring of water quality in shallow aquatic environments. The system integrates spectral measurements with in-situ water color parameter (WCP) collection, offering a more automated and cost-effective alternative to traditional manual methods, and demonstrating its utility in a case study on Poyang Lake.
What Problem Does This Paper Address?
Accurate monitoring of water quality parameters (WQPs) like chlorophyll-a (Chl-a), turbidity, and Colored Dissolved Organic Matter (CDOM) is crucial for understanding and managing aquatic ecosystems. Traditional methods for collecting in-situ spectral data and WCPs are often labor-intensive, time-consuming, and prone to human error, especially in dynamic shallow water environments. While satellite remote sensing offers broad coverage, its temporal resolution can be insufficient for capturing rapid changes. Existing automated systems often lack full integration or are not designed for long-term, stationary deployments. This paper aims to bridge this gap by developing a fully automated, stationary system that can provide continuous, high-frequency, and reliable data for both in-situ monitoring and validation of remote sensing products.
Methods
The core of the AFWCPOS is an Automatic Stationary Water Color Parameters Observation System (AFWCPOS). Key components and methodologies include:
- Automatic Sun-Tracking Platform (ASTP): This platform is designed to automatically adjust the viewing geometry of the sensors to minimize glint and optimize measurements, a critical challenge in above-water spectral measurements. It calculates the solar azimuth angle to maintain consistent measurement conditions.
- Spectral Measurement Unit: Six TriOS RAMSES hyperspectral spectroradiometers are employed. Four are configured for radiance measurements (Lu and Lsky) at specific viewing angles (135° or 45° from nadir), and two are for downwelling irradiance (Ed) measurements. These sensors cover the spectral range from 350 nm to 950 nm.
- Water Color Parameter (WCP) Measurement Unit: Three WETLabs ECO fluorometers are used for in-situ measurement of Chl-a, turbidity, and CDOM. These sensors are deployed underwater and provide continuous data.
- Data Acquisition and Wireless Transmission: A data logger (KAMTLabs ALL-IN-ONE) manages system configuration, data collection, storage, and wireless transmission (via GPRS/CDMA). A GPS device provides location and time stamps.
- Above-Water Measurement Protocol: The system follows the above-water method for spectral reflectance (Rrs) determination, which involves measuring total upwelling radiance (Lu), sky radiance (Lsky), and downwelling irradiance (Ed). The Rrs is calculated by correcting Lu for skylight reflection (Lsky) and accounting for the downwelling irradiance. The ASTP automates the precise viewing geometry required to minimize errors.
Datasets
- Poyang Lake In-situ Data: The AFWCPOS was deployed in Poyang Lake, China, from July 20-28, 2013. During this period, the system collected continuous spectral data and WCPs (Chl-a, turbidity, CDOM).
- SVC HR-1024 Field Spectroradiometer Data: Simultaneous spectral measurements were collected using a portable SVC HR-1024 spectroradiometer from a ship for cross-comparison and validation of the AFWCPOS spectral data.
- Landsat 8 OLI Image: A Landsat 8 OLI image of Poyang Lake was used to demonstrate the remote sensing application of the AFWCPOS data. Empirical relationships were established between the AFWCPOS-derived WCPs and simulated Landsat 8 OLI band reflectance to map spatial distributions.
- Lab-based Fluorometric Analysis: Chl-a concentration data obtained from lab-based fluorometric analysis of water samples were used to validate the in-situ ECO fluorometer measurements.
Key Insights and Ideas
- Automation is Key for Dynamic Waters: The study highlights that manual measurements are impractical for capturing the diurnal and inter-diurnal variations common in high-dynamic shallow waters like Poyang Lake. The AFWCPOS's automation is crucial for obtaining such data.
- Integrated Systems Offer Synergistic Benefits: Combining spectral measurements with direct WCP sensing in a single automated platform provides a powerful tool for both in-situ monitoring and for generating robust datasets for satellite data validation.
- Sun-Tracking Significantly Improves Data Reliability: The ASTP's ability to maintain optimal viewing geometry is critical for minimizing glint and improving the accuracy of Rrs measurements, a persistent challenge in above-water techniques.
- Cost-Effectiveness for High-Frequency Monitoring: The AFWCPOS offers a more cost-effective solution for achieving high-frequency temporal sampling compared to repeated ship-based surveys or extensive manual sampling campaigns.
- Potential for Broad Application: The system's design and successful application in Poyang Lake suggest its applicability to a wide range of shallow coastal and inland waters facing similar monitoring challenges.
Key Results
- Spectral Data Consistency: The AFWCPOS spectral data showed high consistency with measurements from a conventional SVC HR-1024 field spectroradiometer, with a coefficient of determination (R²) of 0.96, an unbiased percent difference (UPD) of 0.14, and a mean relative difference (MRD) of 0.078.
- WCP Validation: In-situ ECO Chl-a measurements showed good agreement with lab-based fluorometric analysis, with a linear relationship exhibiting a slope of 1.05 and an R² of 0.72.
- Diurnal and Inter-diurnal Variations: The system successfully captured significant diurnal and inter-diurnal variations in Chl-a, turbidity, and CDOM in Poyang Lake.
- Remote Sensing Application: Empirical models developed using AFWCPOS data and Landsat 8 OLI bands allowed for the mapping of spatial distributions of Chl-a, turbidity, and CDOM, which were consistent with previous research and reflected known hydrodynamic and nutrient influences in Poyang Lake.
Limitations and Future Work
- Fouling of Optical Sensors: The paper acknowledges the issue of optical sensor fouling in moored deployments, requiring daily manual cleaning.
- Instrument Discrepancies: Some disagreements between AFWCPOS and SVC measurements are attributed to differences in instrument parameters (FOV, SNR), measurement conditions, and manual operation of the SVC.
- Cost Reduction: Future work could focus on reducing system costs by using spectral sensors based on high-speed multiplexers or self-developed sensors.
- Onboard Calibration: The authors plan to incorporate onboard calibration instruments and methods to further enhance data reliability.
- Error Mitigation: Further research will explore methods to eliminate errors arising from different optical sensors and to ensure sensors are not contaminated by atmospheric ash.
- Expanded Field Deployment: Increasing the number of field sensors could allow for more data collection without interference, improving dynamic water quality monitoring.
Summary updated