Street-level solar radiation mapping and patterns profiling using Baidu Street View images

Abstract The variability of urban landscapes and weather conditions makes it challenging to quantify the solar radiation of urban streets. In this study, we propose a framework to map and profile the street-level solar radiation using Baidu Street View (BSV) images. In our framework, a BSV-based radiation estimation method considering weather conditions is established to map the hourly street-level solar radiation. Then, a two-stage clustering method is proposed to profile the radiation patterns, including daily radiation patterns (DRP) and multifaceted radiation patterns (MRP), from the estimated radiation data. Taking Chongqing as a case, results show that (1) the estimated solar radiation of the proposed method is closer to the real situation than the two existing methods. (2) The solar radiation of streets varies greatly among different road grades and streetscapes; the street environment and weather conditions have a significant impact on solar radiation in mountain cities. (3) Eight types of DRP and six types of MRP are discovered by the two-stage clustering method. Moreover, the features, geographical distribution and applications of the radiation patterns are investigated in detail to support urban planning. This work can greatly promote studies relating the solar radiation at the street-level in the future.