Comprehensive Dataset for Data-Driven Pavement Performance Prediction in Flood-Prone Beaumont, Southeast Texas

Timely and effective pavement maintenance is essential for ensuring transportation safety, operational efficiency, and long-term economic sustainability. The evaluation of pavement condition, encompassing parameters such as structural strength, surface roughness, and distress manifestation is critical for guiding infrastructure management decisions. Pavement performance metrics directly influence vehicle handling, ride comfort, and overall roadway safety. In recent years, growing emphasis has been placed on the application of data-driven models to predict pavement deterioration, with the goal of improving Maintenance and Rehabilitation (M&R) planning and optimizing resource allocation.

The success of these predictive approaches depends fundamentally on the availability of standardized, high-quality datasets capable of capturing the multifaceted drivers of pavement degradation. This data article introduces a comprehensive dataset developed to advance research in pavement performance prediction, with a regional focus on Southeast Texas, specifically the flood-prone urban area of Beaumont. The dataset integrates a broad range of variables, including pavement condition and traffic data, meteorological records, synthetic flood model outputs, ground deformation indices, and terrain-based topographic features. Together, these inputs allow for the assessment of both load-related and environmental deterioration mechanisms.

Data processing and integration were conducted using ArcGIS Pro, Microsoft Excel, and Python to ensure consistency and usability for data-driven analytical frameworks, particularly Machine Learning applications. This dataset offers several key contributions: it supports the exploration of environmental and climatic influences on pavement health, aids in identifying critical variables affecting performance, and enables detailed correlation and trend analyses across heterogeneous data types.

By addressing existing limitations in input variable selection for pavement prediction models, the dataset lays the groundwork for more accurate forecasting of maintenance requirements and promotes resilience-oriented planning in areas vulnerable to flooding. Overall, this work underscores the pivotal role of harmonized data resources in enhancing pavement management systems and supports ongoing efforts to develop robust, data-informed strategies for infrastructure preservation.