Real-Time Kinematic (RTK) and Network Real-Time Kinematic (NRTK) are mainstream technologies for real-time centimeter-level positioning based on GNSS. Tropospheric delay, as a major error source in RTK positioning, is particularly significant in medium to long baseline positioning. Due to its close correlation with atmospheric humidity variations, precise quantification of this error is challenging. This study, based on the NRTK method, utilizes GNSS sample data from four reference stations to estimate Zenith Tropospheric Delay (ZTD), ambiguity, and position parameters through double-difference processing. Innovatively, prior interpolated ZTD is introduced as an observation value, effectively enhancing positioning accuracy and Integer Ambiguity Resolution (IAR) performance. The research employs ZTD interpolation data from the IGS reference station network, combined with the Weighted Least Squares method for solution. Experimental results indicate that, compared to traditional RTK methods with unknown ZTD, this technique achieves an IAR success rate of nearly 100%.