Assessing the impact of drought on Zlatar Lake area changes using Random Forest with Sentinel-1 SAR and Sentinel-2 multispectral data
Drought, a complex natural hazard, poses significant challenges for water resource management due to its unpredictable nature. This study investigates the impact of the drought on Zlatar Lake in Serbia, a key artificial reservoir in the Uvac Valley, over a period extending from May 2021 to October 2...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Savez inženjera i tehničara Srbije
2024-01-01
|
| Series: | Tehnika |
| Subjects: | |
| Online Access: | https://scindeks-clanci.ceon.rs/data/pdf/0040-2176/2024/0040-21762406663N.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832096606333173760 |
|---|---|
| author | Nikolić Ratko R. |
| author_facet | Nikolić Ratko R. |
| author_sort | Nikolić Ratko R. |
| collection | DOAJ |
| description | Drought, a complex natural hazard, poses significant challenges for water resource management due to its unpredictable nature. This study investigates the impact of the drought on Zlatar Lake in Serbia, a key artificial reservoir in the Uvac Valley, over a period extending from May 2021 to October 2022. By utilizing Sentinel-1 Synthetic Aperture Radar (SAR) and Sentinel-2 Multispectral Instrument (MSI) imagery, and classifying 17 scenes, the research employs the Random Forest classification algorithm to analyze changes in the lake's water area. Key indices, including the VV/VH ratio, Normalized Difference Polarization Index (NDPI), and various spectral water indices (NDWI, MNDWI, SWI, NDVI), were used to enhance classification results. The study achieved high overall accuracy (OA) and Kappa values in classifying individual scenes, with Sentinel-1 reaching an OA of 96.47% and a Kappa coefficient of 92.71, while Sentinel-2 showed an OA of 99.14% and a Kappa coefficient of 98.20. Results revealed a significant water area decline, with Sentinel-1 data decreasing from 7.12 km² in May 2021 to a minimum of 4.60 km² in June 2022, then rising to 5.22 km² by October 2022. Sentinel-2 data showed a similar trend. This study highlights the effectiveness of combining SAR and optical data for monitoring water body changes and provides valuable insights for drought impact assessment and water resource management. |
| format | Article |
| id | doaj-art-a68bf3ace89e4387b17eabb6945c8f47 |
| institution | Kabale University |
| issn | 0040-2176 2560-3086 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | Savez inženjera i tehničara Srbije |
| record_format | Article |
| series | Tehnika |
| spelling | doaj-art-a68bf3ace89e4387b17eabb6945c8f472025-02-05T13:30:15ZengSavez inženjera i tehničara SrbijeTehnika0040-21762560-30862024-01-0179666367210.5937/tehnika2406663N0040-21762406663NAssessing the impact of drought on Zlatar Lake area changes using Random Forest with Sentinel-1 SAR and Sentinel-2 multispectral dataNikolić Ratko R.0https://orcid.org/0009-0000-5964-030XVekom Geo d.o.o., Belgrade, SerbiaDrought, a complex natural hazard, poses significant challenges for water resource management due to its unpredictable nature. This study investigates the impact of the drought on Zlatar Lake in Serbia, a key artificial reservoir in the Uvac Valley, over a period extending from May 2021 to October 2022. By utilizing Sentinel-1 Synthetic Aperture Radar (SAR) and Sentinel-2 Multispectral Instrument (MSI) imagery, and classifying 17 scenes, the research employs the Random Forest classification algorithm to analyze changes in the lake's water area. Key indices, including the VV/VH ratio, Normalized Difference Polarization Index (NDPI), and various spectral water indices (NDWI, MNDWI, SWI, NDVI), were used to enhance classification results. The study achieved high overall accuracy (OA) and Kappa values in classifying individual scenes, with Sentinel-1 reaching an OA of 96.47% and a Kappa coefficient of 92.71, while Sentinel-2 showed an OA of 99.14% and a Kappa coefficient of 98.20. Results revealed a significant water area decline, with Sentinel-1 data decreasing from 7.12 km² in May 2021 to a minimum of 4.60 km² in June 2022, then rising to 5.22 km² by October 2022. Sentinel-2 data showed a similar trend. This study highlights the effectiveness of combining SAR and optical data for monitoring water body changes and provides valuable insights for drought impact assessment and water resource management.https://scindeks-clanci.ceon.rs/data/pdf/0040-2176/2024/0040-21762406663N.pdfdroughtsentinel - 1 sarsentinel - 2 msirandom forest |
| spellingShingle | Nikolić Ratko R. Assessing the impact of drought on Zlatar Lake area changes using Random Forest with Sentinel-1 SAR and Sentinel-2 multispectral data Tehnika drought sentinel - 1 sar sentinel - 2 msi random forest |
| title | Assessing the impact of drought on Zlatar Lake area changes using Random Forest with Sentinel-1 SAR and Sentinel-2 multispectral data |
| title_full | Assessing the impact of drought on Zlatar Lake area changes using Random Forest with Sentinel-1 SAR and Sentinel-2 multispectral data |
| title_fullStr | Assessing the impact of drought on Zlatar Lake area changes using Random Forest with Sentinel-1 SAR and Sentinel-2 multispectral data |
| title_full_unstemmed | Assessing the impact of drought on Zlatar Lake area changes using Random Forest with Sentinel-1 SAR and Sentinel-2 multispectral data |
| title_short | Assessing the impact of drought on Zlatar Lake area changes using Random Forest with Sentinel-1 SAR and Sentinel-2 multispectral data |
| title_sort | assessing the impact of drought on zlatar lake area changes using random forest with sentinel 1 sar and sentinel 2 multispectral data |
| topic | drought sentinel - 1 sar sentinel - 2 msi random forest |
| url | https://scindeks-clanci.ceon.rs/data/pdf/0040-2176/2024/0040-21762406663N.pdf |
| work_keys_str_mv | AT nikolicratkor assessingtheimpactofdroughtonzlatarlakeareachangesusingrandomforestwithsentinel1sarandsentinel2multispectraldata |