Modified Priority-Flood algorithm for hydrologic modelling-based digital terrain analysis using a hash heap structure

Digital elevation models (DEMs) employed in hydrological simulations require preprocessing to mitigate the adverse effects of depressions, which represents a foundational challenge in physics-based hydrological modelling. Although various methods exist for processing DEM depressions, the growing volume of DEM data imposes increasing demands on the computational efficiency of existing algorithms. Given that computational performance is a critical criterion for method selection among spatial hydrology practitioners, this study built upon the fastest monolithic Priority-Flood algorithm and introduced an optimization of its priority queue mechanism using a hybrid data structure (Hash Heap (HHeap)) designed to address the identified efficiency bottlenecks in the queue. HHeap integrates the advantages of hash tables and heaps to optimize queue performance in geospatial computations. The proposed architecture features a dual-component design that achieves two key improvements: (1) a substantial reduction in structural reorganization frequency caused by repetitive terrain patterns commonly encountered in DEMs and (2) the preservation of optimal time complexity for core operations (O(1) for queries and O(log n) for insertions and deletions). Empirical evaluations across multiscale DEM datasets (ranging from 108 to 1010 cells) demonstrated that the enhanced algorithm achieved computational speedups of 2–25% (mean: 10%) relative to the fastest monolithic Priority-Flood algorithm without increasing memory consumption. Since the HHeap structure optimizes the original priority queue, it significantly enhances the computational performance of the Priority-Flood algorithm while maintaining complete consistency with existing Priority-Flood variants in terms of accuracy. This feature provides an efficient and reliable computational foundation for next-generation terrain analysis systems.