In recent decade, there has been an abundant generation of high-dimensional data in larger volumes while measuring the water resource environment. This environment consists of large spatial areas with high resolution and high temporal data, where such collection needs to be summarized. In this paper, we develop a deep learning model to recognize the meaningful relationship between the objects in data during decision-making process. Appropriate extraction of meaningful relationship using deep learning reduces the redundant information that provides inter-variable relationships and characteristic patterns. The deep learning framework recognizes the patterns via clustering the data that provides better understanding of objects. The experimental results are conducted with various training options and parameter selection on the information extracted. The results of simulation shows that the proposed deep learning model achieves improved clustering than other methods.