Spatiotemporal dynamics of chlorophyll-a in the Gorgan Bay and Miankaleh Peninsula biosphere reserve: Call for action

Remote Sensing Applications Society and Environment, Journal Year: 2023, Volume and Issue: 30, P. 100946 - 100946

Published: March 2, 2023

Language: Английский

---

B1D–CNN: A Novel Convolution Neural Network-Based Chlorophyll-A Retrieval Algorithm for Sentinel–2 Data

IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium, Journal Year: 2023, Volume and Issue: unknown, P. 3950 - 3953

Published: July 16, 2023

Deep learning (DL) methods have been recently considered suitable for Chlorophyll-a (Chla) retrieval from satellite data due to their ability handling complex, high-dimensional, and noisy data. This manuscript describes B1D–CNN, a new model combining 1D-convolutional neural networks (1D-CNN) traditional empirical blend algorithm estimate Chla using the MultiSpectral Instrument (MSI) sensor on board Sentinel-2 satellite. The proposed is trained evaluated against state-of-the-art Mixture Density Network (MDN) classical algorithms global in-situ results show 9.25% 54.12% improvement in RMSE, along with 3.45% 59.53% reduction MAE. A image during harmful algal bloom (HABs) event was also assessed captured high batches associated HABs. study indicates advantages of DL retrieve Chla.

Language: Английский

---

1D Convolutional Neural Network-based Chlorophyll-a Retrieval Algorithm for Sentinel-2 MultiSpectral Instrument in Various Trophic States

Sensors and Materials, Journal Year: 2023, Volume and Issue: 35(11), P. 3743 - 3743

Published: Nov. 28, 2023

Despite extensive research on chlorophyll-a (Chla) concentration retrieval methods from remote sensing reflectance (Rrs, sr -1 ) data, there remains a need for more reliable Chla techniques.In this study, we introduce deep learning approach based 1D convolutional neural network (1D CNN) architecture.In addition, provide new method of representing the Rrs as sequential vector.The model architecture targets Sentinel-2 MultiSpectral Instrument (MSI) sensor.The proposed was trained and tested simulated in situ data collected broad trophic states Japan Vietnam waters with concentrations ranging 0.02 to 148.26 mg/m 3 .The evaluated against well-accepted state-of-the-art methods: ocean color three-band (OC3), index (OCI), two-band ratio, Blend, mixture density network.The evaluation shows that outperforms other 7.48-38.02%reduction root mean squared error (RMSE) an 11.50-39.17%lower absolute (MAE) than methods.The promising performance suggests attention should be paid domain sequence modeling CNN.

Language: Английский

---

Data-Driven Interpolation of Sea Surface Suspended Concentrations Derived from Ocean Colour Remote Sensing Data

Remote Sensing, Journal Year: 2021, Volume and Issue: 13(17), P. 3537 - 3537

Published: Sept. 6, 2021

Due to complex natural and anthropogenic interconnected forcings, the dynamics of suspended sediments within ocean water column remains difficult understand monitor. Numerical models still lack capabilities account for variabilities depicted by in situ satellite-derived datasets. Besides, irregular space-time sampling associated with satellite sensors make crucial development efficient interpolation methods. Optimal Interpolation (OI) state-of-the-art approach most operational products. large increase both measurements more available information is coming from measurements, as well simulation models. The emergence data-driven schemes possibly relevant alternatives increased recover finer-scale processes. In this study, we investigate benchmark three schemes, namely an EOF-based technique, analog data assimilation scheme, a neural network approach, OI scheme. We rely on Observing System Simulation Experiment based high-resolution numerical simulations simulated observations using real patterns. which relies variational formulation problem, clearly outperforms other terms reconstruction performance greater ability high-frequency events. further discuss how these results could transfer data, problems beyond issues, especially short-term forecasting partial observations.

Language: Английский

---

Estimation Method of Chlorophyll Concentration Distribution Based on UAV Aerial Images Considering Turbid Water Distribution in a Reservoir

Drones, Journal Year: 2024, Volume and Issue: 8(6), P. 224 - 224

Published: May 29, 2024

The observation of the phytoplankton distribution with a high spatiotemporal resolution is necessary to track nutrient sources that cause algal blooms and understand their behavior in response hydraulic phenomena. Photography from UAVs, which has an excellent temporal spatial resolution, effective method obtain water quality information comprehensively. In this study, we attempted develop for estimating chlorophyll concentration aerial images using machine learning considers brightness correction based on insolation turbidity evaluated by satellite image analysis. reflectance harmful algae bloom (HAB) was different seen under normal conditions; so, containing HAB were causes error estimation concentration. First, when occurred extracted discrimination learning. Then, other used regression Finally, coefficient determination between estimated no analysis observed value reached 0.84. proposed enables detailed depiction concentration, contributes improvement management reservoirs.

Language: Английский

---