Solar3D: A 3D Extension of GRASS GIS r.sun for Estimating Solar Radiation in Urban Environments

Abstract

Solar3D is an open-source software application designed to interactively calculate solar irradiation at three-dimensional (3D) surfaces in a virtual environment constructed with combinations of 3D city models, digital elevation models (DEMs), digital surface models (DSMs) and feature layers. The GRASS GIS r.sun solar radiation model computes solar irradiation based on two-dimensional (2D) raster maps for given day, latitude, surface and atmospheric conditions. With the increasing availability of 3D city models and demand for solar energy, there is an urgent need for better tools to computes solar radiation directly with 3D city models. Solar3D extends GRASS GIS r.sun from 2D to 3D by feeding the model with input, including surface slope, aspect and time-resolved shading, that is derived directly from the 3D scene using computer graphics techniques. To summarize, Solar3D offers several new features which, as a whole, distinguish itself from existing 3D solar irradiation tools: (1) the ability to consume massive heterogeneous 3D city models, including massive 3D city models such as oblique airborne photogrammetry-based 3D city models (OAP3Ds or integrated meshes); (2) the ability to perform near real-time pointwise calculation for duration from daily to annual; (3) the ability to integrate and interactively explore large-scale heterogeneous geospatial data. (4) the ability to calculate solar irradiation at arbitrary surface positions including at rooftops, facades and the ground. Solar3D is publicly available at https://github.com/jian9695/Solar3D.

SolarGAN: Synthetic Annual Solar Irradiance Time Series on Urban Building Facades via Deep Generative Networks

“…In addition, • Regarding the acquisition of data, especially the way we capture fisheye categorical mask images, although this is not the focus of this paper, using ClimateStudio to render such simple images is an obvious overkill, which significantly impedes the overall speed. In the future, alternative approaches similar to [7,8], which are based on taking real-time screenshots in the GIS platform, need to be investigated. Even so, reading, writing, and transferring images might be inevitably rather time-consuming in practice.…”

“…Compared to CAD software with powerful capabilities of 3D operations, GIS platform-based simulation methods are often limited to 2D. To consider the influence of urban context, several recent studies have introduced methods based on the classical Sky View Factor (SVF) or shadow mask [6,7,8]. Requiring only low computational cost operations such as projection and screenshot, these methods cleverly obtain 2D circular fisheye binary masks to represent the skydome of a sensorpoint and reflect the 3D shadowing effect of the surrounding urban context.…”

Stochastic Solar Irradiance from Deep Generative Networks and their Application in BIPV Design

“…These time-series samples are stochastically sampled conditioned on given location, date, and weather condition, etc.. In addition, inspired by conventional simulation approaches based on shadow mask [4] and sky view factor [16,17], we realize that a circular fisheye projection image can be an excellent form to represent the sky visibility and urban context conditions in the sunlit hemisphere of a test point on any building surface. Furthermore, such images can be easily captured from open-source urban LOD1 geometry databases.…”

“…The acquisition of data, especially the way to capture the fisheye categorical mask images, although this is not the focus of this paper, can be significantly streamlined. In the future, alternative approaches similar to [16,17], which are based on taking real-time screenshots from multiple directions on a GIS platform, need to be investigated. Even so, reading, writing, and transferring images can be timeconsuming in practice.…”