In this study, a new empirical model for the relationship between Peak Ground Acceleration (PGA) and Arias Intensity is proposed, based on strong-motion database from shallow crustal earthquakes in Iran.

First, a comprehensive database was compiled, including event date and time, epicentral location, earthquake magnitude, distance metrics, and site characteristics of the recording stations (soil classification or shear-wave velocity, where available). Subsequently, 3,117 high-quality three-component records were selected from the strong-motion network of the Building and Housing Research Center (BHRC). The selected records underwent baseline correction and filtering; following processing, PGA and Arias Intensity values were computed for all records.

Based on this database, various functional forms were examined to represent the relationship between Arias Intensity and PGA. Through residual analysis and evaluation of standard deviations, the most suitable empirical model was identified. To enrich the database in the near-source region, a set of global records within comparable distance and magnitude ranges was incorporated into the Iranian dataset. The empirical relationships were then re-developed using this combined database (Iranian records supplemented with global near-source, high-magnitude records).

The results of this study indicate that the relationships derived from the combined database exhibit greater accuracy and stability. Their application is recommended for engineering purposes and seismic hazard assessments for earthquakes with moment magnitudes ranging from 3.5 to 7.8, recorded at Joyner-Boore distances up to 200 km. Furthermore, quantitative evaluations demonstrate that the inclusion of near-fault data resulted in a 3% reduction in the overall standard deviation. Notably, comparisons with established global models indicate good agreement with the proposed model in this study; specifically, when considering magnitudes between 5.0 and 7.0 recorded on various soil classes, the discrepancies range from 5% to 33%.