Mesoscale unfrozen water content measurements in frozen oasis soils using combined cosmic-ray neutron sensing and time domain reflectometry networks

Frozen soil water is a key indicator for assessing soil thermal regimes, moisture status, and freeze-thaw dynamics, which are crucial for understanding soil hydrological processes in oasis ecosystems. However, accurately measuring mesoscale unfrozen and frozen water contents during freeze-thaw cycles remains challenging. This study combines Cosmic-Ray Neutron Sensing (CRNS) and Time Domain Reflectometry (TDR) networks to investigate ice-water transformation characteristics at the mesoscale in oasis ecosystems, including farmland with shallow groundwater (FSG), old farmland (OF), new farmland (NF), and artificial forest (AF) during the freezing period. The results show that CRNS effectively measured mesoscale soil moisture with high accuracy in desert-oasis ecoregion (R² = 0.930, RMSE = 0.027 cm 3 /cm 3 ). While, the parameter N 0 for farmland sites was similar, though it was lower for AF. The integrated CRNS and TDR networks effectively differentiated the frozen water content from the total value during the freezing period at the mesoscale. Unfrozen water content ranged from 0.04 to 0.48 cm³ /cm³ , with more variation in surface layers compared to deeper layers. Freezing occurred from the top to down in OF, NF, and AF, whereas FSG exhibited a bidirectional freezing pattern. Freezing rates were highest in FSG, followed by NF, AF, and OF. The OF retained a higher proportion of unfrozen water content, while AF maintained the lowest ratio throughout the freezing period. Soil temperature was the dominant factor. Below 0°C, unfrozen water content exhibited an exponential relationship with temperature, while above 0°C, it followed a power relationship. These findings enhance our understanding of freeze-thaw processes and provide valuable insights for monitoring mesoscale unfrozen soil moisture in oasis ecosystems. • CRNS and TDR networks effectively captured mesoscale soil freezing dynamics. • Landscape heterogeneity significantly influenced freezing rates and ice-water dynamics. • Unfrozen soil water was mainly controlled by soil temperature, following exponential and power-law patterns.