Abstract: With the rapid development of satellite observation technology, relatively accurate estimates of the energy budget at the top of the earth’s atmosphere have become achievable. However, at the earth’s surface, the estimation error in the energy balance is still formidable. Among the various components contributing to the energy imbalance estimation at the earth’s surface, downward shortwave solar radiation, termed as surface solar radiation (SSR) herein,represents one of the most important sources of error. SSR is not only important for simulating land surface processes,but also serves as a key indicator for the utilization of renewable solar photovoltaic energy. Therefore, accurateobservation of SSR is crutial for surface energy balance calculation and related applications. At the same time, long-term variations of SSR have always been a major concern across various fields. This review starts with a comparison of existing SSR observational products, discusses the bias and uncertainty issues in the long-term variations of SSR,and clarifies the importance of developing high-quality SSR baseline data products. Then, the present paper introduces a series of studies in recent years, which, based on the most complete SSR station data to date, systematically examined the inhomogeneity and sampling problems, performed systematic homogenization processing and artificial intelligence (AI) reconstruction on station series, and estimated the long-term variations and uncertainty levels of SSRs at global and regional scales. These studies offer new evidence for global and regional climate change observation, detection, attribution, and future projection. Finally, the paper presents an outlook on the existing and future challenges in the research on SSR data and the SSR long-term variations.