Solar Photovoltaics (PV) performance is influenced by several environmental factors, including humidity, temperature, shading, and soiling. Out of these, soiling significantly degrades PV power performance and the effective lifetime of solar panels. Various studies, including outdoor, controlled indoor environments and soiling models, have been conducted to understand the soiling phenomenon and devise effective mitigation strategies. This review evaluates the impact of various factors influencing soiling loss from experimental through to simulation studies. Through a critical analysis of existing literature focused on publications till November 2025, various research gaps have been identified. These gaps include the need for comparative and comprehensive outdoor soiling studies that consider a greater scope of influencing parameters, such as installation height, smoke, smog, and bird droppings, for a range of PV technologies in different climates. The need for greater development of controlled indoor soiling chambers is identified. Current set-ups lack smoke, smog, rainfall control, wind turbulence, and wind gusts, which are key soiling influencing parameters, as well as the system being able to accommodate regular-sized PV panels and ideally PV arrays. The reported soiling models are site-specific and highly dependent on outdoor experimental data, where the reliability of results is uncertain. The development of these models also does not consider all the important parameters that will impact soiling. Based on these research gaps, this review offers recommendations for future research direction regarding outdoor studies, advancements in indoor soiling chambers, and accurate estimation of soiling loss through modelling. • A comprehensive overview of global aerosol cycle, detailing its generation sources, transportation, and adhesion processes. • An in-depth review of technical and climatic factors affecting solar PV soiling. • A controlled environment is important to single out the impact of various parameters. • Machine and deep learning models are more suitable for soiling loss estimation due to the complex and non-linear nature of soiling data. • Hybrid models based on experimental and controlled environment studies can help to estimate the soiling loss more accurately.
Hussain et al. (Sun,) studied this question.