This study presents a probabilistic analysis of geomagnetically induced Global Positioning System (GPS) loss of lock (LoL) at the low-latitude ZOMB station during Solar Cycle 24. Daily and component-wise LoL statistics reveal that most disruptions occur at the signal level, with satellite- and epoch-level losses being progressively rarer. Seasonal and monthly fractional analyses indicate mid-year peaks in signal instability, reflecting ionospheric variability characteristic of low-latitude, scintillation-prone regions. Conditional probability assessments demonstrate that higher AE index, F10.7 solar flux, and RMS By magnetic field fluctuations systematically increase the likelihood of LoL events. Predictor diagnostics and principal component analysis identify a compact set of environmental variables that mitigate multicollinearity and provide robust inputs for probabilistic modelling. Multiscale wavelet analysis further characterises temporal variability, showing contributions from short-term, seasonal, and multi-year cycles. This integrated methodology, combining component-wise characterisation, space weather linkage, and multiscale temporal analysis, offers novel insights into GPS vulnerability to space weather and provides a framework for operational monitoring and mitigation strategies.