This paper describes the features of the multidimensional spatial effects among G20 financial markets and investigates the dynamic evolution process of financial impact in the spatial spillover networks. To achieve this objective, we first define the economic metric and gravitational spatial weight matrix to establish a multidimensional economic space, which is shown to have a significant advantage in capturing the multidimensional spatial effects of financial markets. Then, the spatial econometric complex network autoregressive panel model is constructed by combining the spatial econometrics and complex network theory. Finally, we calculate the Bonacich centrality of nodes in spillover network, and study dynamic process of the financial impact path with different Bonacich nodes being the source of infection. The results show that spatial spillover network includes regional aggregation and dynamic evolution. The spatial spillover effects of financial markets are positively correlated with important financial markets, foreign direct investment, turnover ratio and inflation rate. Emerging financial markets are seen to have increasingly become Bonacich key nodes in the impact path. When the Bonacich key nodes in the spillover network have fluctuations, the impact spreads wider and faster. This study can help us predict the source of the financial turmoil, and remind investors to formulate new asset allocation strategies and risk prevention measures for different financial markets.