This dataset is a global monthly drought data set with a spatial resolution of 0.25 ° from 1948 to 2010. It is based on a multi type, multi-scale drought index, namely the Standardized Humidity Anomaly Index (SZIsnow) that includes snow dynamics, and is driven by the system field of advanced data assimilation systems. The proposed SZIsnow dataset includes different physical water energy processes, especially snow processes. Our evaluation of this dataset indicates that it can distinguish different drought types at different time scales. Our evaluation also indicates that the dataset is capable of fully capturing drought at different spatial scales. The consideration of snow accumulation processes has improved the ability of SZIsnow, especially in high latitude areas covered by snow (such as the Arctic region) and high-altitude areas (such as the Qinghai Tibet Plateau). We found that between 1948 and 2010, 59.66% of the Earth's land area showed a dry trend, while the remaining 40.34% showed a wet trend. Our results also indicate that the SZIsnow dataset can be used to capture large-scale drought events occurring around the world. Our analysis shows that during the study period, there were 525 drought events with an area greater than 500000 square kilometers worldwide, of which 68.38% lasted for more than 6 months. Therefore, this new drought dataset is highly suitable for monitoring, evaluating, and describing drought, and can serve as a valuable resource for future drought research.