Fine mapping of surface water plays an important role in water resource management, ecological environment protection, and flood prevention and control. However, the Songliao River Basin still lacks high-precision, long-term surface water distribution datasets. To fill this gap, this article constructs a surface water extraction algorithm suitable for the Songliao River Basin based on multi-source satellite remote sensing data (Landsat 5/7/8/9) and the Google Earth Engine (GEE) platform, generating an annual surface water distribution dataset from 1984 to 2023. This dataset includes annual water body frequency data (WBF) and water body classification data (SWC), covering seasonal water bodies, permanent water bodies, and maximum water body distribution. Through precision evaluation, the overall accuracy of the water extraction method reached 98%, with a Kappa coefficient of 0.95. In order to verify the reliability of the data, a comparative analysis was conducted between the SWC data and the Global Water Classification History (GWS) dataset. The results showed that the classification accuracy of the SWC dataset in the Songliao Basin was better than that of the GWS dataset in multiple regions. The high-precision water body dataset provided in this study will provide important support for water resource monitoring and management in the Songliao River Basin, and will also lay a solid data foundation for scientific research in related fields
This dataset covers annual surface water frequency (WBF) and surface water classification (SWC) data for the Songliao River Basin in Northeast China from 1984 to 2023. SWC data contains category information for permanent water bodies, seasonal water bodies, and maximum water bodies. The data is presented in raster format using the WGS-84 geographic coordinate system, with a spatial resolution of 30 meters and a file format of *. tif. The naming convention is "YYYY. tif" (where YYYY represents the corresponding year). The band value range of surface water frequency data (WBF) is 0 to 1, which represents the floating point value of water observation frequency and is used to reflect the frequency of water occurrence in different years. The band values of surface water classification data (SWC) are 0 and 1, where 0 represents non water bodies and 1 represents water bodies, used to distinguish the spatial distribution of water bodies and non water bodies. This dataset provides accurate spatiotemporal data support for surface water dynamic monitoring and water resource management in the Songliao River Basin, and can play an important role in hydrology, ecology, and environmental protection
(1) Using the GEE tool, cloud removal and preprocessing were performed on Landsat-5/7/8/9 data, with a projection coordinate system of WGS84
(2) Extract water bodies from Landsat-5/7/8/9 images
(3) Use HAND terrain index to correct water data and evaluate the accuracy of water extraction methods
(4) Calculate the dataset of water frequency and water category
The data quality is good
| # | number | name | type |
| 1 | 242102321160 | Henan Provincial Science and Technology Research Project |
This work is licensed under a
Creative
Commons Attribution 4.0 International License.
CSTR:11738.11.NCDC.CSD.DB6687.2024
10.12072/ncdc.csd.db6687.2024