In a groundbreaking advancement for climate-based daylight modeling, researchers Aw, Leng, and Lim have developed an innovative Principal Component Analysis (PCA)-based climatic similarity index designed specifically to enhance weather file selection criteria in tropical climates. This pioneering study, published in Scientific Reports in 2026, not only addresses the critical challenge of accurately simulating daylight conditions in regions characterized by complex tropical weather patterns but also marks a significant leap forward in climate-driven architectural and environmental modeling.
Tropical climates are notoriously difficult to model due to their inherent variability and a rich tapestry of meteorological phenomena, including high humidity, fluctuating cloud cover, intense rainfall, and dynamic solar radiation patterns. Traditional weather file selection methods often fall short in capturing this complexity, which can lead to inaccuracies when used as inputs for climate-based daylight simulations. It is precisely this gap that the PCA-based climatic similarity index aims to bridge, by offering a sophisticated, data-driven method to select representative weather files that better mirror the actual climatic nuances of tropical regions.
At the core of this new method lies Principal Component Analysis, a powerful statistical tool that reduces the dimensionality of large, multivariate datasets while preserving the most critical variances within the data. By applying PCA to a wide array of climatic variables—such as temperature, humidity, solar radiation, wind speed, and cloud cover—the researchers extracted patterns that effectively distill the complexity of tropical weather into manageable, representative components. This dimensionality reduction facilitates the identification of weather days or periods that are climatically similar, which can then be selectively employed in daylight modeling simulations to enhance their fidelity.
One of the most striking aspects of this research is its focus on tailoring the index specifically for tropical environments. Unlike temperate or arid regions, tropical climates present a unique set of challenges due to their lack of distinct seasons, diverse microclimates, and frequent convective weather events. The PCA-based similarity index developed in this study accommodates these factors by incorporating variables sensitive to the high variability and transient nature of tropical weather, thereby offering a bespoke solution rather than a one-size-fits-all approach.
Implementing this index drastically improves the selection of weather files for daylight modeling. Instead of relying on standardized Typical Meteorological Year (TMY) datasets, which might not sufficiently capture the local climatic variability, the index enables the identification of the most climatologically representative data points. Consequently, simulations built on these carefully curated datasets demonstrate markedly improved accuracy in predicting daylight performance, solar heat gains, and glare potential—parameters crucial for the optimization of building energy efficiency, occupant comfort, and environmental impact in tropical architecture.
The implications of this development extend beyond just daylight modeling. Because accurate weather file selection is foundational for a host of building performance simulations, the PCA-based climatic similarity index holds promise for applications in thermal comfort analyses, energy consumption forecasting, and even in the design of adaptive building envelopes that respond dynamically to changing climatic conditions. By improving the input data quality, the model ensures that downstream simulation outputs are more reliable and relevant to real-world scenarios.
Technically, the process begins by assembling an extensive local weather dataset comprising high-resolution meteorological measurements collected over multiple years. The researchers then preprocess the data to address missing values and normalize the variables, ensuring that the subsequent PCA accurately reflects genuine climatic patterns rather than artifacts or biases. The PCA is executed to derive principal components that quantify the dominant modes of variability in the dataset. Each weather day is then scored based on these components, facilitating clustering or similarity assessments.
The novel similarity index is computed by measuring the Euclidean distance between weather days in the principal component space, enabling the quantification of climatic resemblance. Days with the smallest distances are deemed most similar, allowing researchers and practitioners to construct weather file subsets that authentically capture the spectrum of climatic conditions typical for the region. This approach reveals weather analogues that traditional chronological or statistical selection methods might overlook.
To validate the performance of their index, the authors integrated it within a daylighting simulation framework and compared simulation outcomes using weather files selected via their method against those obtained from standard TMY and random selections. The results showed improved correlation with observed daylighting metrics, reduced simulation uncertainty, and an enhanced capacity to capture extreme weather events that significantly influence daylight availability and building performance in the tropics.
This methodological innovation also paves the way for more robust scenario-based simulations in the context of climate change. As tropical regions face increasing variability due to global warming, having a dynamic, adaptable index that can recalibrate weather file selection in response to shifting climatic baselines becomes essential. The PCA-based similarity measure offers a flexible tool to explore the impacts of future climate scenarios on daylight and building performance, supporting resilience and adaptation strategies.
Moreover, this approach is transferable to other climate-based modeling domains. While the current work focuses on daylighting, the underlying mathematical framework could be repurposed for selecting representative weather data in hydrological models, urban microclimate assessments, and even in renewable energy generation forecasting where tropical climate variability plays a critical role.
The researchers emphasize the importance of open data and algorithm transparency, ensuring that the index can be implemented in open-source simulation tools frequently used by the building performance community. This accessibility enhances the potential for widespread adoption, community-driven refinement, and integration into international building standards addressing tropical architecture.
In conclusion, Aw, Leng, and Lim’s development of a PCA-based climatic similarity index represents a vital advancement for climate-sensitive modeling in tropical regions. By refining the selection of weather files, their work promises to improve the accuracy and reliability of daylight simulations, contributing to more sustainable, comfortable, and energy-efficient building designs. As tropical cities grow rapidly and strive to meet sustainability goals, such innovative tools become invaluable for architects, engineers, and policymakers navigating the complexities of climate-responsive design.
Subject of Research: Development of a Principal Component Analysis-based climatic similarity index for enhanced weather file selection in climate-based daylight modeling in tropical climates.
Article Title: Development of a PCA-based climatic similarity index to enhance weather file selection criteria for climate-based daylight modelling simulations in tropical climates.
Article References:
Aw, S.B., Leng, P.C. & Lim, Y.W. Development of a PCA-based climatic similarity index to enhance weather file selection criteria for climate-based daylight modelling simulations in tropical climates. Sci Rep (2026). https://doi.org/10.1038/s41598-026-58112-x
Image Credits: AI Generated
