A risk assessment of GNSS PNT service degradation is essential for the establishment of quality of services (QoS) of GNSS-enabled applications. Space weather, geomagnetic and ionospheric conditions were found to be the prime class of the GNSS positioning performance degradation causes. In this paper, we address the problem of provision of insight to risks of the GNSS PNT deployment for GNSS-based applications with a long-term risk analysis of GPS positioning accuracy degradation and its effects on GNSS air, maritime, and Location-Based Services (LBS) applications in tropical regions. We assembled a large database of single-frequency GPS positioning error vector estimates derived from experimental stationary GPS pseudorange observations at the International GNSS Service (IGS) reference station at Darwin, NT, and append it with the Kobe University hourly Dst index of geomagnetic activity estimates. Further, we analysed the statistical properties of the GPS positioning error vectors in classes of geomagnetic conditions, determined with Dst index values. The extent of single-frequency non-aided GPS positioning accuracy cases not meeting the recognised requirements for GNSS air, maritime and Location-Based Services is identified. Furthermore, we find of a number of cases in which the smoothed GPS positioning error vector exceeded the limits acceptable for GNSS air, maritime, and LBS applications, serving as outliers. The expected risk based on the long-term GPS positioning error vector observations is calculated from the number of outlier cases of selected GNSS air, maritime, and LBS applications in selected ranges of Dst index values. This allows for the risk estimation of GPS positioning performance not meeting the requirements of GNSS applications requirements in specific geomagnetic conditions, as determined with Dst values. Finally, we study the other potential GPS positioning performance disruption candidate descriptors, validate, and identify those that may serve in the GNSS deployment risk assessment for targeted GNSS-based applications.