Monitoring a Short-Term Dry-to-Wet Soil-Moisture Transition Using GPS Multipath in Tropical Environment

Abstract

Continuous local observations of surface soil moisture remain limited in humid tropical regions, where optical monitoring is frequently interrupted by cloud cover. This study evaluates whether an operational Continuously Operating Reference Station can resolve a short dry-to-wet transition using Global Navigation Satellite System Interferometric Reflectometry. GPS L2C signal-to-noise ratio observations from the CPLU station in North Padang Lawas, North Sumatra, were processed at elevation angles of 10°–25° and azimuths of 0°–180°. Direct-signal trends were removed, and accepted multipath arcs were screened using Lomb–Scargle peak criteria. Track-specific phase changes were transformed into a consistently scaled daily relative volumetric water content series. A stable dry interval identified from SMAP supplied the lower relative baseline, and September 2025 was retained as the focused transition month. GNSS-IR-derived relative volumetric water content (VWC-GNSSIR) increased from 0.070 m³/m³ at the beginning of September to 0.290 m³/m³ near month-end. The monthly mean was 0.179 m³/m³, and local rainfall totaled 135.6 mm across 17 rainy days. Across six cloud-free Sentinel-2 observations from the broader season, the original NDVI showed a strong descriptive association with VWC-GNSSIR (r = 0.871), whereas a three-month moving composite produced a weaker relationship (r approximately 0.390). The results demonstrate that CORS-based GNSS-IR can capture short-term tropical soil-moisture dynamics that are substantially smoothed in temporally aggregated optical products.