Large-scale integration of wind power into grid has significant impact on system operation and stability due to the inherent intermittency of wind power. Consequently

accurate prediction of regional wind farm output has become one of the vital challenges for the power industry. However

most of the existing prediction methods only involve mathematical models and rarely consider spatial correlation factors. As such

the prediction system remains to be further improved. Wind power prediction based on spatial correlation is a comprehensive prediction method that combines both local information and spatially correlated information. This paper introduces the definition

concepts and basic characteristics of the wind power prediction based on spatial correlation. Specific aspects such as statistical model

physical model

down-scaling and up-scaling procedure in terms of wind power prediction based on spatial correlation are also explained in detail. The state-of-the-art of spatial correlation approaches and techniques used in wind power prediction around the world are systematically reviewed and summarized. Finally

the future development trend and further investigations are presented for the existing problems and shortcomings in this field.