SVR model and OLCI images reveal a declining trend in phycocyanin levels in typical lakes across Northeast China
Recently, the frequency of cyanobacterial blooms in lakes across Northeast China has noticeably increased, because of the combined effect of global climate change and human activities. Phycocyanin (PC), a distinctive pigment found in cyanobacteria, plays a pivotal role in detecting and predicting ha...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | Ecological Informatics |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1574954124005077 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Recently, the frequency of cyanobacterial blooms in lakes across Northeast China has noticeably increased, because of the combined effect of global climate change and human activities. Phycocyanin (PC), a distinctive pigment found in cyanobacteria, plays a pivotal role in detecting and predicting harmful cyanobacterial blooms. A novel segmented PC concentration remote sensing inversion model was developed using the support vector regression (SVR) algorithm and Sentinel-3 OLCI data to examine the spatiotemporal variations in PC concentrations across 11 representative lakes in Northeast China. The model showed high accuracy (PC ≤ 45 μg/L: R2 = 0.882, MAPE = 34.53 %, MAE = 3.96 μg/L; PC > 45 μg/L: R2 = 0.814, MAPE = 31.80 %, MAE = 70.58 μg/L). Significant variations in PC concentrations were observed across the watersheds in Northeast China. Over seven years, the PC concentrations declined in half of the lakes. Although no significant changes in the annual average PC concentrations were noticed, intra-annual fluctuations were observed. Seasonal patterns in PC concentrations were also observed. The contribution rates of natural and anthropogenic factors were determined using spatiotemporal data on PC concentrations from 2016 to 2022 and redundancy analysis. The primary drivers of PC concentrations varied across different spatial and temporal scales. These findings highlight the monitoring of PC concentrations in the lakes of Northeast China using Sentinel-3 OLCI images. This study can provide valuable insights for water quality management authorities, enabling them to understand the changing characteristics and trends of PC concentrations and to issue early warnings of cyanobacterial blooms. |
|---|---|
| ISSN: | 1574-9541 |