Air pollution remains a major environmental and health challenge in Indonesia, driven by rapid urbanization, industrial expansion, and transport emissions. This study provides a systematic review of the progress, challenges, and future directions in the performance and application of air-quality models across Indonesia from 2010 to 2024. A total of 122 peer-reviewed studies were analyzed using PRISMA 2020 guidelines, covering deterministic models such as AERMOD, CALINE4, WRF-Chem, CALPUFF, and HYSPLIT, as well as emerging machine-learning approaches. Results show that deterministic models remain dominant for urban and industrial assessments, yet their performance is limited by incomplete emission inventories, sparse monitoring networks, and complex tropical meteorology. Recent advances using machine learning, low-cost sensors, and satellite data have improved forecasting, though integration with policy and regulatory frameworks remains limited. Overall, Indonesia’s modeling landscape is progressing but fragmented. Strengthening emission databases, enhancing model validation, and improving collaboration between research institutions and policymakers are essential, providing key scientific evidence to support the development of data-driven and policy-integrated air quality management frameworks in tropical archipelagic environments. • Review of 122 air-quality modeling studies in Indonesia (2010–2024) • Deterministic models dominate but face data and validation limitations • Machine learning shows strong predictive performance • Hybrid modeling offers a promising future direction • Policy integration remains a key challenge
Bachtiar et al. (Wed,) studied this question.