Water Frequency and Water Change Class dataset (2017-2024)

This is the dataset of this article "Mapping small water bodies and their dynamics with deep learning: fusion of Sentinel-1 and Sentinel-2 imagery at semi-monthly 10 m spatial resolution". This dataset comprises annual water frequency data for Hubei Province from 2017 to 2024, along with water change classification data over the eight-year period from 2017 to 2024. The values of the water frequency data range from 0 to 1, with higher values indicating a higher observed frequency of water. The water change classification data consists of 9 categories, each represented by a distinct value, where -2 indicates invalid values, 0 represents land, 1 represents permanent water, 2 represents stable seasonal water, 3 represents disappearing water (loss), 4 represents newly formed water (gain), 5 represents water with a decreasing then increasing trajectory (dry period), 6 represents water with an increasing then decreasing trajectory (wet period), and 7 represents highly dynamic water (high frequency).