Surface water is an essential component of water resources. Obtaining spatial information on surface water quickly and accurately is of great significance for the management and utilization of water resources and the protection of the ecological environment. The research on water extraction methods using remote sensing images is a trending topic that has been extensively explored in water index, classification, subpixel analysis, and other related areas. Compared with other methods, a water-index-based method has the advantages of speed and convenience. The characteristics of surface water, such as its extensive coverage and instability, make the water index particularly effective in monitoring large areas of surface water. We propose a new Landsat-8 image water index combination (WI2023). We also utilize the ratio of the near-infrared band and short-wave infrared band to eliminate the shadow between water and non-water, thereby enhancing the contrast between water and other ground objects. According to the different terrains, the performance of WI2023 was verified in plain, hill, mountain, plateau, and basin areas and compared with four widely used water indexes. The results show that WI2023 has obvious advantages in water extraction in hilly areas compared with the other four methods (i.e., normalized differential water index, modified normalized differential water index, fringe cap humidity index, and multi-band water index). The mean overall accuracy and kappa coefficient are higher at 96.16% and 0.94, respectively. WI2023 holds the potential to be a useful surface water extraction technology for water resource studies and applications.