An Estimation Method on Thermophysical Properties of the Building Surface Based on Multispectral Remote Sensing and Surface Energy Balance Simulation

Surface energy balance (SEB) models, exploring the interaction between buildings and the atmosphere at the surface layer, can identify areas with severe outdoor thermal conditions, examine potential technologies for the UHI mitigation, and can be coupled with building energy models for a more accurate energy demand prediction. The selection of thermophysical properties is important to the simulation quality. In addition to a study that assumes standardized building parameters, collecting the input building material data in the real world is a tedious job because of the large number of diverse materials used in urban constructions. Few studies have derived the surface covering materials and albedo by remote sensing, but the thermal properties (e.g., heat capacity and thermal conductivity) of urban surfaces are usually assumed and generalized. The current methods (e.g. the theoretical method, in-situ method, and etc.) that most widely used to obtain thermal properties of walls cannot be well adopted on an urban block or neighborhood.

To bridge this gap, a new time- and labor-saving estimation method is proposed in this study to obtain the actual thermophysical properties, coupling the use of multi-spectral remote sensing and numerical simulation. This estimation method can shorten the measurement period, extent the study scale (e.g., block and neighborhood), and reduce some requirements of the current in-situ methods for assessing the thermal transmittance of walls, such as the many pieces of equipment, strict operation, steady weather condition, certain times of the day. A case study is thus carried out on a typical summer day with clear sky conditions in a city block in Yokohama city, Japan. On that day, hourly infrared and multi-spectral images for target wall surfaces were captured on ground from sunrise to sunset and around midnight. A SEB model (THERMORender) and input parameters (meteorological data and physical properties) are introduced and validated. The input albedos of the target wall surfaces were estimated by using the multi-spectral images and validated by the accurate spectrum measured by using a spectroradiometer. The input thermal properties of 10 simulation cases were sourced from a database of commonly used building surface materials in an urban area of Japan, characterizing the typical thermal property values and their combination. The simulated hourly radiant temperatures for the target wall surfaces of 10 cases were compared to the corresponding in-situ measured infrared data by a pixel-to-pixel comparison. The thermal properties in the case which result showed most consistent to the measured infrared data were the ones closest to reproduce the radiant and heat exchange between the building surface and the atmosphere among 10 typical cases, which can be regarded as the estimated values of the actual thermal properties of the target wall surfaces. The estimated values of thermal properties obtained by our method here were validated by a theoretical method based on ISO 6946, with the known composition and thermal properties of layers for the target wall surfaces. The validation demonstrated that the estimation results are representative because the estimated and theoretical values are less than 20% different.

Therefore, the performance of the estimation method proposed in this study is sufficient. This estimation method is limited to the prior known typical combinations of thermal properties of building surfaces. To integrate with the vehicle or unmanned aerial vehicle, this estimation method is allowed for obtaining the actual thermophysical properties of urban surfaces in order to improve the SEB model performance on an urban block or neighborhood.