Predictive maintenance is essential in aviation due to rising cost pressures, leveraging sensor data and maintenance logs for anticipatory maintenance, improving planning efficiency. Analyzing historical data ensures timely interventions, reducing unplanned downtime and enhancing aircraft reliability. Digital twin integration expands these capabilities, allowing precise monitoring and predictive analyses of aircraft components, tracking stress, fatigue, and predicting conditions. Accurate load monitoring during ground operations requires integrating actual aircraft trajectories with environmental factors like pavement conditions and weather, which can pose challenges due to data sparsity, noise, or misalignment. Our study outlines a methodology using sparse ADS-B and geospatial airport data, employing map-matching and filtering techniques for comprehensive trajectory representation and analysis. Additionally, we introduce roughnes-specific pavement modeling stochastically to allow precise load assessment on aircraft structures during ground roll including surface variations and damage patterns. This model precedes a probabilistic fatigue model, aiming to diagnose, predict, and mitigate potential structural issues, thereby enhancing aircraft durability and operational safety.