Dust impact on photovoltaics: a systematic review of meteorological and environmental soiling prediction models

This study reviewed literature on PV soiling loss estimation to understand the viability of forecasting PV system performance penalties imposed by dust using environmental and meteorological modeling. Dust dramatically decreases PV module transmittance in arid/semi-arid climates; monthly average energy losses have been reported to exceed 30% and values up to 80% during soiling intervals if modules are not cleaned periodically. Revenue losses and increased operation and maintenance budgets are incurred as a result. Scopus, Web of Science, and Google Scholar databases were searched for literature from Jan. 2000 to Mar. 2026 using PRISMA keywords, resulting in 59 eligible articles contributing 41 distinct, parameterized soiling prediction models. Models were reviewed and categorized by the predictors used to estimate normalized soiling rate, such as particulate concentration, wind speed, relative humidity, rainfall, days since last cleaning (DSLSTC), and dust properties. Literature-derived models were grouped into 5 categories: cleanness-index models, empirical regression models, particulate-based models, meteorological-only models, and machine-learning models. Reported performance metrics (coefficient of determination, R 2 ; root mean squared error, RMSE; mean absolute percentage error, MAPE) were then compared amongst model families. Purely particulate- or machine-learning-based models often had the highest accuracies within calibration sites, with some R 2 values exceeding 0.85 and MAPE between 3 and 5%. Meteorological- and cleanness-index models had lower accuracies but demonstrated broader transferability between sites. The article recommends a standard suite of soiling performance metrics, highlights gaps in multi-climate validation and hybrid-model development at large temporal and spatial scales, and presents model-selection guidance based on review findings. We hope this work helps PV plant operators and research scientists better understand and predict soiling in dusty climates.