From SWAT to Hybrid SWAT+: A Conceptual Review of the Evolution of Watershed Modelling towards AI-Enhanced Environmental Informatics

Abstract

The escalating complexity of global water resource challenges driven by climate change, land-use transformation, and intensifying anthropogenic pressures has reinforced the strategic importance of physically based hydrological models as critical decision-support tools. This review synthesises the conceptual evolution of the Soil and Water Assessment Tool (SWAT), tracing its transition from a subbasin-centric, file-based architecture to the modular, object-oriented environmental informatics framework of SWAT+. Central to this evolution is the replacement of proprietary MS Access databases with open-source SQLite and the introduction of Landscape Units (LSUs), which improve spatial representation by organising Hydrologic Response Units (HRUs) within geomorphologically meaningful landscape hierarchies. Global evaluations indicate that SWAT+ generally provides improved hydrological performance over legacy SWAT, particularly in representing spatial connectivity and baseflow processes, although challenges remain in small headwater catchments and arid regions. Beyond reviewing this technological evolution, the principal contribution of this article is the development of a conceptual Hybrid SWAT+ Modelling Framework. The proposed framework integrates the physically based modelling core with artificial intelligence and real-time environmental data across three complementary levels: pre-processing for input enhancement, in-simulation via rule-based decision tables, and post-processing for residual correction. By addressing parameter non-stationarity while preserving mass balance and process interpretability, this hybridisation enables adaptive learning under changing environmental conditions. The proposed framework provides a conceptual pathway toward adaptive watershed modelling and next-generation environmental informatics, supporting more climate-resilient and evidence-based water resources management.