This thesis investigates the influence of atmospheric effects, primarily the troposphere, on Global Navigation Satellite Systems (GNSS) positioning accuracy and evaluates methods for mitigating these effects in precise positioning applications. The work focuses on the modeling of tropospheric delay, including the use of different mapping functions, horizontal gradient modeling, and height correction of gridded zenith delays. In addition, the potential of GNSS-derived tropospheric delays for estimating atmospheric water vapor is explored. The performance of three mapping functions, Minimum Operational Performance Standards (MOPS), Niell Mapping Function (NMF), and Vienna Mapping Function 3 (VMF3), is evaluated using a dual approach consisting of numerical ray tracing through numerical weather models and observation-based Precise Point Positioning (PPP) processing using the Ginan software. The ray-tracing framework enables controlled analysis of slant delay modeling and the contribution of horizontal gradients, while the PPP processing reflects real-world performance through residual and positioning accuracy. The results show that advanced mapping functions significantly improve the stability of tropospheric modeling compared to simpler models, with VMF3 providing the most consistent performance. However, the difference between VMF3 and NMF is generally small, indicating diminishing returns with increased model complexity in many practical scenarios. In contrast, the inclusion of horizontal gradient terms has a substantial impact, particularly at low elevation angles and during dynamic atmospheric conditions. Furthermore, the analysis demonstrates that proper height correction of gridded zenith delays is critical for achieving accurate results, in some cases having a larger impact than the choice of mapping function itself. The study also shows that low-cost GNSS receivers are capable of capturing the general temporal behavior of integrated water vapor, suggesting potential for dense networks of inexpensive sensors in meteorological applications. Overall, the findings highlight the importance of a balanced modeling strategy, where gradient modeling and height correction play a central role alongside the choice of mapping function.