A new algorithm for mapping large inland water bodies using CYGNSS

ABSTRACT

The persistence of inland water bodies affects climate, biodiversity, and human societies. Among the multiple remote sensing methods, Global Navigation Satellite System Reflectometry (GNSS-R) technology has shown great potential for inland water bodies mapping at very high temporal resolutions. For large inland water bodies with dimensions larger than 350 km, long fetches would roughen water surfaces, thereby causing low-power GNSS returns, posing challenges for mapping them using GNSS-R data, particularly in the case of ultra-large lakes with complicated shorelines, such as Lake Victoria. In this study, we propose an algorithm for mapping Lake Victoria at a 0.01° × 0.01° spatial resolution using GNSS-R data from the Cyclone Global Navigation Satellite System (CYGNSS). By mainly leveraging the surface reflectivity signal, our algorithm extracts lake boundaries and fills the interior to map Lake Victoria. The probability of detection (POD) of the resultant water mask was approximately 90%. This simple and robust algorithm could enhance the capability of monitoring global fast-changing inland water bodies using GNSS-R data, especially in the pan-tropical areas.

KEYWORDS:

• GNSS-R
• large inland water bodies
• Lake Victoria
• CYGNSS

Acknowledgement

This study was supported by the Natural Science Foundation of China (No. 42041005-4). The CYGNSS data are freely available at https://cygnss.engin.umich.edu/data-products/. We thank the Google Earth Engine (GEE) platform, which provides publicly accessible data and computing power. The authors are grateful to the anonymous reviewers and the editor for their constructive and excellent reviews, which greatly improved the quality of the article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Credit author statement

Maoxiang Chang, and Peng Li: Writing, Investigation; Maoxiang Chang, Peng Li, Zhenhong Li: Conceptualization, Supervision; Peng Li, Zhenhong Li, and Maoxiang Chang: Resources; Peng Li, Zhenhong Li, and Maoxiang Chang: Software; Maoxiang Chang, Yue Sun, and Peng Li: Formal analysis, Data Curation; Maoxiang Chang, Peng Li, and Zhenhong Li: Methodology; Maoxiang Chang, Peng Li, Yufeng Hu, Yue Sun, Zhenhong Li, and Houjie Wang: Review, Validation.

Data availability statement

The example code for CYGNSS data processing and large inland water bodies mapping algorithms are openly available through Zenodo (https://zenodo.org/record/7832537). Other materials related to this study are available from the corresponding author upon reasonable request.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [42041005-4].