Spatiotemporal variability and driving factors of the shallow soil moisture in North China during the past 31 years

To fully understand the dynamic change characteristics of soil moisture in North China, the spatiotemporal patterns of soil moisture anomalies (SMA) in the four active soil layers from January 1990 to December 2020 are analyzed based on the Global Land Data Assimilation System (GLDAS) model, then the driving factors and corresponding contribution rates (CRs) for SMA change are determined together with the meteorological, socioeconomic and land cover data. The results show that the SMA in the last three layers present decreasing trends, and the trend magnitudes increase in multiples with the increase of soil depth. The area with the most severe soil dry is clustered in the east-central plain, while the relatively wet areas are scattered in western Shanxi and northern Hebei. The annual, intra- and inter-annual cycles (1.0, 0.5, 4.3 and 7.3 yr) are the dominant frequency of the SMA time series, which are driven jointly by 6 natural-type factors and 6 human-type factors, especially the intensified evaporation and dramatically expanding urbanization. By combining principal component analysis (PCA) and grey absolute correlation degree, the reintegrated human and natural factors respectively account for 49.46% and 50.54% of mean CRs to SMA in the four layers, but the human factor dominates the change of the bottom SMA. On the other hand, the determined mean CRs by combining exploratory factor analysis (EFA) and grey absolute correlation degree are 53.46% and 46.44% for reintegrated positive and negative factors, but the negative factor has a greater effect on the bottom SMA. The results provide a clear clarification of the soil moisture dynamic change, which is conducive to shedding new light on water resources management, agricultural security and ecological sustainability.