The task of predicting the influence of scientific research and ranking authors is a substantial difficulty that has attracted the attention of scientists in different fields of study. This task is crucial for improving research efficiency, supporting decision-making, and enabling scientific evaluations. Various indexes have been proposed by the scientific community to identify significant authors, including citation count, publication count, hybrid techniques, the h-index, and its variants. However, the scientific community does not agree on the most suitable single index for author ranking. This study presents a new index derived from a dataset in the field of Civil Engineering. The dataset consists of information from 500 authors, with an equal distribution between those who have not received awards and those who have received awards from prestigious civil engineering scientific societies during year 2011 to 2019. Initially, we ranked indices based on their values for each individual author to find the top parameters regularly placing awardees in the top 10% and 20% records. We also utilized Random Forest techniques to identify the top influential parameters. Subsequently, the correlation between the h-index and its various derivatives was calculated. A strong correlation suggested that these indices had a similar effect, while a weak correlation implied little or no relationship between them. The study leads us to suggest a new index derived by aggregating the top two parameters, so that it outperforms by 10% in terms of accuracy relative to the indices evaluated in this study. The awardees honored by different civil engineering societies serve as a benchmark for evaluating the impact of authors in this study. The awardees occurring in each index’s ranked list are determined, and the extent to which each index contributed to bringing awardees to the top is evaluated. It has been shown that the PI index is capable of placing up to accuracy 60% and 69% awardees positioned in the top 10% and 20% of ranked list respectively.