Monitoring pelagic Sargassum in the Atlantic Ocean from space: Principles and practices

Harmful Algae, Journal Year: 2025, Volume and Issue: unknown, P. 102840 - 102840

Published: March 1, 2025

Language: Английский

---

Significant Improvement in Short-Term Green-Tide Transport Predictions Using the XGBoost Model

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(9), P. 1636 - 1636

Published: May 5, 2025

Accurately predicting the drift trajectory of green tides is crucial for assessing potential risks and implementing effective countermeasures. This paper proposes a short-term green-tide prediction method that combines patch characteristics, 1 h interval distances from GOCI-II images, driving-factor data using XGBoost machine learning model to enhance accuracy. The results demonstrate proposed outperforms traditional OpenDrift in predictions. Specifically, at time intervals 3, 5, 7 h, root mean square errors (RMSEs) zonal direction are 1.81 km, 2.89 3.55 respectively, whereas RMSEs 0.80 0.98 1.20 respectively; meridional direction, 1.77 2.67 3.10 while 0.82 1.10 1.25 respectively. Furthermore, more-accurately tracks actual positions patches compared model. 25 interval, continues accurately predict positions, exhibits significant deviations. study demonstrates method, by patterns historical data, effectively predicts process tides. It provides valuable support early warning systems, thereby helping mitigate ecological economic impacts disasters.

Language: Английский

---

Microalgae proliferation and influences of water to feed stock ratio on hydrogen generation from Ulva Prolifera through super critical water gasification process

International Journal of Hydrogen Energy, Journal Year: 2025, Volume and Issue: 136, P. 31 - 39

Published: May 9, 2025

Language: Английский

---

A Novel Approach of Monitoring Ulva pertusa Green Tide on the Basis of UAV and Deep Learning

Water, Journal Year: 2023, Volume and Issue: 15(17), P. 3080 - 3080

Published: Aug. 28, 2023

Ulva pertusa (U. pertusa) is a benthic macroalgae in submerged conditions, and it relatively difficult to monitor with the remote sensing approaches for floating macroalgae. In this work, novel remote-sensing approach proposed monitoring U. green tide, which applies deep learning method high-resolution RGB images acquired unmanned aerial vehicle (UAV). The results of extraction from semi-simultaneous UAV, Landsat-8, Gaofen-1 (GF-1) demonstrate superior accuracy extracting UAV images, achieving an 96.46%, precision 94.84%, recall 92.42%, F1 score 0.92, surpassing algae index-based method. also performs well satellite 85.11%, 74.05%, 96.44%, 0.83. cross-validation between Landsat-8 root mean square error (RMSE) portion (POM) model 0.15, relative difference (MRD) 25.01%. POM reduces MRD area imagery 36.08% 6%. This combining tends enable automated, high-precision pertusa, overcoming limitations approach, calibrate image-based improve frequency by applying when are not available.

Language: Английский

---

Unprecedent green macroalgae bloom: mechanism and implication to disaster prediction and prevention

International Journal of Digital Earth, Journal Year: 2023, Volume and Issue: 16(1), P. 3772 - 3793

Published: Sept. 18, 2023

Green macroalgae bloom (GMB), with the dominant species of Ulva prolifera, has regularly occurred since 2007 along China coast. Although disaster prevention and control achieved favorable results in 2020, satellite-observed GMB annual maximum coverage (AMC) rebounded sharply 2021 to an unprecedented level. The reasons for this rebound significant interannual variability over past 15 years are still open questions. Here, by using long-term time-series (2007–2022) optical Synthetic Aperture Radar satellite observations (1000+ scenes), meteorological data water quality statistics, mechanism analysis was performed exploring effects from natural factors human activities. Two key determinants AMC successfully identified numerous potential which distribution a area (the Subei Shoal) during critical period (from April May 20) nutrient availability. Furthermore, these two parameters, novel model prediction (R2 = 0.87, p < 0.01) is proposed independently validated, can reasonably explain (2014–2021) agree well latest observation 2022 (percentage difference 12%). Finally, suggestions future alleviation. This work may aid management measure optimization.

Language: Английский

---