Multi-scale Modelling of Potential Solar Power Generation in Urban Areas
Decentralized power generation through photovoltaics (PVs) is widely regarded as contributing favorably to environmental, economic, and social aspects of urban sustainability. The optimized use of decentralized solar PV power generation, however, requires estimation of the potential for building surfaces and their electricity demand, as well as the design of smart distribution systems and transmission grids. Here we propose methods to estimate the solar PV power generation in urban areas in an efficient way. The methods apply to individual buildings, neighborhoods, cities as well as entire countries. Our results show that GIS and LiDAR point data-based methods are promising approaches for estimating the solar PV power generation on building roofs and facades at the scales from individual buildings to cities. LiDAR methods are, however, difficult to apply to entire countries (national scale), in which case Machine-Learning methods are more appropriate. Generally, whenever massive amount of data are available, as for whole countries, Machine-Learning methods are a good alternative to LiDAR-based methods for the estimation of solar PV power generation on building roofs and facades.
Rooftop photovoltaics, spatio-temporal mapping, solar-power generation, cities