2. Materials and Methods
In this paper, we used CiteSpace, a Java application for visualizing and analyzing trends and patterns in scientific literatures, for scientometric analysis. The paper data used for the analysis was extracted from the Science Citation Index-Expanded (SCI-E) of the Web of Science Core Collection (WoSCC). Specific retrieval strategy is as follows: hydrological connectivity OR hydrological connection OR hydrologic connection OR hydrologic connectivity. Also, timespan was set as 1998-2018. This search obtained 3,307 bibliographic records including titles, abstracts, and cited references and were then exported to CiteSpace for subsequent analysis. A knowledge map of the publication characteristics, including subject categories, journals, keywords, authors and references was generated based on a certain number of papers published in this field using the functions of Citespace software (co-cited analysis, co-occurrence analysis and co-operative analysis), which shows the hot research issues and emerging trends. A network map contains many nodes and links. Nodes in different kinds of network map represent different meaning such as author, keyword, reference, institution, and the larger the nodes, the more the publications or frequency (Liang, Luo & Zhong, 2018). The connecting lines between nodes indicate a relationship of co-operation, co-occurrence or co-citation, and their various colors represent the year that the two nodes are co-cited (Chen, 2006; Xie, 2015; Min et al., 2018). Additionally, a node surrounded by a purple ring indicates that the node has a relatively strong (≥ 0.1) centrality, which indicated an extensive connection with other nodes. In this research, most of the analysis was conducted by CiteSpace based on papers published on SCI-E (Science Citation Index Expanded) over the period of 1998 to 2018. Contributing co-cited authors and journals were mapped to identify the evolution of hydrological connectivity. Hot issues and frontier research were identified based on the frequency of popular keywords, which can be divided into 3 periods (i.e., 1998 to 2004, 2005 to 2011, and 2012 to 2018). Analysis of highly co-cited references will also be conducted as an effective supplement to the related research hotspots. In addition, burst detection of co-cited authors was applied to identify the author whose research achievements in this research field is significant. Therefore, the analysis of co-operative and co-occurrence network as well as co-cited ones with the frequency, centrality and burst will be conducted to identify research characteristics and emerging trends on hydrological connectivity.