Development of a new spatially complete and daily continuous lake surface water temperature dataset for Lake Vänern by fusing satellite and reanalysis products
(2025) In Journal of Hydrology: Regional Studies 59.- Abstract
Study region: Lake Vänern, the largest lake in the European Union, located in Sweden. Study focus: The study aimed to develop a spatially complete and temporally continuous lake surface water temperature (LSWT) dataset by fusing satellite remote sensing data with reanalysis products. Five satellite-derived LSWT products, including the Moderate Resolution Imaging Spectroradiometer (MODIS), were evaluated against in-situ measurements. The reanalysis product ERA5-Land, offering hourly LSWT at a spatial resolution of 0.1°, was fused with the MODIS LSWT product using the Enhanced Spatial and Temporal Adaptive Reflectance Model (ESTARFM). ESTARFM combines high spatial but low temporal resolution data with products of low spatial but high... (More)
Study region: Lake Vänern, the largest lake in the European Union, located in Sweden. Study focus: The study aimed to develop a spatially complete and temporally continuous lake surface water temperature (LSWT) dataset by fusing satellite remote sensing data with reanalysis products. Five satellite-derived LSWT products, including the Moderate Resolution Imaging Spectroradiometer (MODIS), were evaluated against in-situ measurements. The reanalysis product ERA5-Land, offering hourly LSWT at a spatial resolution of 0.1°, was fused with the MODIS LSWT product using the Enhanced Spatial and Temporal Adaptive Reflectance Model (ESTARFM). ESTARFM combines high spatial but low temporal resolution data with products of low spatial but high temporal resolution. New hydrological insights for the region: This study demonstrates the effectiveness of data fusion in generating accurate and continuous LSWT datasets for large lakes. The fused LSWT product achieved a mean absolute error of 1.53 °C and R² of 0.86, showing comparable accuracy to ERA5-Land data. While a bias-correction approach was tested, results indicated that seasonal adjustments may be more effective. The fused dataset captured spatial and temporal variations in LSWT, aligning well with in-situ measurements and existing datasets. This approach enables improved LSWT monitoring and provides a valuable tool for studying ecological changes in lakes driven by climate change. The method can be applied to other large lakes with suitable datasets, supporting regional and global hydrological studies.
(Less)
- author
- Schultze, Anna LU ; Berggren, Martin LU and Duan, Zheng LU
- organization
- publishing date
- 2025-06
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Data fusion, Inland water, Lake surface water temperature, Reanalysis product, Remote sensing
- in
- Journal of Hydrology: Regional Studies
- volume
- 59
- article number
- 102437
- publisher
- Elsevier
- external identifiers
-
- scopus:105004771296
- ISSN
- 2214-5818
- DOI
- 10.1016/j.ejrh.2025.102437
- language
- English
- LU publication?
- yes
- id
- 56e90303-7d35-4de0-8bb5-c1f24e0e009b
- date added to LUP
- 2025-07-30 11:55:40
- date last changed
- 2025-07-30 11:57:02
@article{56e90303-7d35-4de0-8bb5-c1f24e0e009b,
abstract = {{<p>Study region: Lake Vänern, the largest lake in the European Union, located in Sweden. Study focus: The study aimed to develop a spatially complete and temporally continuous lake surface water temperature (LSWT) dataset by fusing satellite remote sensing data with reanalysis products. Five satellite-derived LSWT products, including the Moderate Resolution Imaging Spectroradiometer (MODIS), were evaluated against in-situ measurements. The reanalysis product ERA5-Land, offering hourly LSWT at a spatial resolution of 0.1°, was fused with the MODIS LSWT product using the Enhanced Spatial and Temporal Adaptive Reflectance Model (ESTARFM). ESTARFM combines high spatial but low temporal resolution data with products of low spatial but high temporal resolution. New hydrological insights for the region: This study demonstrates the effectiveness of data fusion in generating accurate and continuous LSWT datasets for large lakes. The fused LSWT product achieved a mean absolute error of 1.53 °C and R² of 0.86, showing comparable accuracy to ERA5-Land data. While a bias-correction approach was tested, results indicated that seasonal adjustments may be more effective. The fused dataset captured spatial and temporal variations in LSWT, aligning well with in-situ measurements and existing datasets. This approach enables improved LSWT monitoring and provides a valuable tool for studying ecological changes in lakes driven by climate change. The method can be applied to other large lakes with suitable datasets, supporting regional and global hydrological studies.</p>}},
author = {{Schultze, Anna and Berggren, Martin and Duan, Zheng}},
issn = {{2214-5818}},
keywords = {{Data fusion; Inland water; Lake surface water temperature; Reanalysis product; Remote sensing}},
language = {{eng}},
publisher = {{Elsevier}},
series = {{Journal of Hydrology: Regional Studies}},
title = {{Development of a new spatially complete and daily continuous lake surface water temperature dataset for Lake Vänern by fusing satellite and reanalysis products}},
url = {{http://dx.doi.org/10.1016/j.ejrh.2025.102437}},
doi = {{10.1016/j.ejrh.2025.102437}},
volume = {{59}},
year = {{2025}},
}