Dynamic evaluation and spatiotemporal evolution of China's industrial water use efficiency considering undesirable output.

Water is a strategic and basic resource for industrial development. The efficient use of water resources is of great significance for the sustainable development of the economy and society. Dynamic SBM model could overcome the shortcomings of static models and reflect inter-temporal efficiency levels. The kernel density curve is used to fit the distribution pattern of industrial water use efficiency and describe its dynamic evolution. Empirical results show that from 2013 to 2017, under the meta-frontier, the industrial water use efficiency values of Beijing, Tianjin, Shandong, Inner Mongolia, and Shaanxi are all 1, and industrial water use efficiency is high, while the industrial water use efficiency values of Sichuan, Guizhou, Anhui, and other provinces are below 0.3, reflecting the low industrial water use efficiency. From 2013 to 2015, China's industrial water use efficiency generally shows a downward trend but begins to rise in the next 2 years. The kernel density curve generally exhibits a bimodal distribution trend and evolves from a "spike shape" to a "broad peak shape".