3254 reservoirs in China provide a total of 512 attributes. These attributes are provided in the upstream catchment area of the reservoir, including:
1) There are two types of upstream catchment areas for reservoirs, namely the complete catchment area and the intermediate catchment area,
2) Watershed topography,
3) Human activities,
4) Climate,
5) Land cover and utilization,
6) Soil and geology.
We also generated time series of 15 meteorological variables. A large dataset on reservoir storage anomalies (92% of data from 3254 reservoirs), water levels (20%), evaporation (98%), and water bodies (99%) is also provided from multiple sources of satellites - radar/laser satellite altimeters+satellite images from Landsat+Sentinel-2. Verifiation_figures records all validation, time series of numbers, and statistical indicators of water level, water area, and storage changes.
Multi source satellites - radar/laser satellite altimeter, Landsat, Sentinel-2 satellite images.
Reservoir water level data: Extract height data from GeoDAR reservoir shapefile; Select height data with water volume greater than 10% using global surface water detectors; The MAD method (median of absolute deviation) was used to remove outliers passed each time; Removed outliers with heights exceeding DEM from altimeter data; Build a time series using the R package "tsHydro".
Reservoir water meter area data: Calculate NDWI (Normalized Difference Water Index); Sampling the appearance of water along the edge to eliminate erroneously detected water; Cut the amount of water generated to a certain value and combine it with the water surface cover to obtain the final water surface cover; Use a quantization based time outlier filter to eliminate errors in reservoir waters.
Reservoir watershed boundary data: Use automatic outlet relocation algorithm to automatically delineate a large watershed quantity.
Watershed horizontal characteristic data: perform necessary data format conversions on relevant attribute datasets; Calculate catchment area features from raster and vector data using different methods; For continuous variables such as temperature and altitude, calculate the mean, maximum, minimum, and range values.
The data quality is good.
This work is licensed under a
Creative
Commons Attribution 4.0 International License.
CSTR:11738.11.NCDC.ZENODO.DB6437.2024
10.5281/zenodo.7664489