Daily Water-Body Extractions and Forecasts for Kazakhstan from Sentinel-3, 2020-2024

A dataset derived from Sentinel-3 satellite imagery provides daily large-scale surface-water monitoring across Kazakhstan from 2020 to 2024. It includes structured point-set representations of water bodies and basin-level next-day forecasts generated using a multi-factor deep learning model incorporating meteorological variables and temporal patterns. The dataset was created by Xiande Wu and shared under a CC-BY-4.0 license.

Use Cases

• Training water-body forecasting models based on the described meteorological variables and temporal memory descriptors.
• Validating remote sensing water extraction methods against the provided Sentinel-2 reference data mentioned in the description.
• Analyzing seasonal and long-term recurrent patterns of water distribution based on the derived daily water series.
• Benchmarking new deep learning architectures for geospatial time-series prediction against the described framework's performance.

Strengths

• Covers a 5-year time range from 2020 to 2024, providing a substantial temporal series.
• Focuses on a specific, large-scale geography (Kazakhstan) for consistent regional analysis.
• Framework performance was validated against Sentinel-2 references and outperformed multiple baselines in experiments.
• Data is structured as point sets to support rapid large-area mapping, as described.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for specific modeling tasks.
• The 13.0 MB file size suggests the dataset may be a processed summary or model outputs rather than raw imagery.

Provenance

Source

figshare, author Xiande Wu.

Collection Method

Derived from Sentinel-3 sensor data using an integrated framework combining atmospheric correction, pixel-level quality control, and spectral-spatial water identification.

Time Range

2020 to 2024

Freshness

Last updated 2026-05 06 07:07:31; freshness should be verified.

Geography

Kazakhstan