Abstract: Earth observation and navigation technologies provide precise understanding of the Earth's system on temporal and spatial dimensions, and the human societal actions reshape the physical world by the transformation of this understanding into new quality productive forces. A new concept and its extensions of spatiotemporal embodied intelligence is defined as the advanced stage of artificial intelligence development and a new fusion direction of BeiDou satellite navigation system (BDS) and remote sensing. Spatiotemporal embodied intelligence refers to a physical intelligent agent in the real world that utilizes high-precision navigation, positioning, and timing, high-resolution remote sensing, and artificial intelligence methods. It can perceive natural environmental elements and spatiotemporal characteristics of human society with human-like precision through its body. It achieves accurate cognition, intelligent decision-making, and scientific manipulation of the natural environment, forming a closed loop of perception-thinking-action. By continuous autonomous optimization and learning, it progressively enhances its intelligence. As an extension, BeiDou spatiotemporal embodied intelligence model is a physical entity developed by humans in the real world, based on the integration of BDS and high-resolution remote sensing technology. By integrating geographic spatiotemporal intelligence with embodied intelligence, it can make intelligent decision-making, precise operation, and coordinated execution control on natural environments and human society objects. It can also achieve the good goal of right time, right place, right object, right degree, and right amount in operating, influencing and transforming real-world physical objects. Furthermore, a spatiotemporal embodied intelligence architecture model based on the intelligent integration of BDS and high-resolution remote sensing is constructed. And the fusion method is proposed, which includes seven layers,i.e. signal, data, information, knowledge, decision-making, execution, and learning. The feasibility and value of spatiotemporal embodied intelligence theory and technology is verified by two case studies of unmanned farming and marine ranch applications in Shandong Province, China. This study can enhance the high-precision perception and scientific understanding of the spatiotemporal heterogeneity and the interaction mechanisms of satellite signals and environmental elements. It is an important new fusion direction of BDS and high-resolution remote sensing .