**2. Methodology**

In order to conduct this study, a traditional approach based on co-occurrence was selected, which included the assessment of productivity, quality, and structural indicators. In this approach, first, the agents with the highest number of publications were identified, and second, the impact of the publications of these authors was analyzed. This type of analysis, particularly with respect to journals, is highly interesting for researchers, given that it constitutes a way to assess the relevance of the journals in which authors publish their studies [39]. Finally, we used mapping techniques to analyze the structure of the network between different agents. The Scopus database was used to select the sample of studies to analyze. This database has proven to be the most suitable for our area of study, enabling us to ensure the selection of a representative sample of the studies carried out on sustainable irrigation (SI). Furthermore, it is easy to access, allows the visualization and analysis of data, and allows data to be downloaded in different formats for subsequent processing using software applications [40]. Nevertheless, if some works on SI are not indexed in the Scopus database, they have not been considered in our sample.

The term used to carry out the search was "sustainable irrigation", and this selection was based on previous studies on the same topic [41–43]. This term was searched for under authors' keywords and titles. The study period selected was 1999 to 2018. Research activity in this topic peaked during these years. Furthermore, this period immediately followed the 1st World Water Forum held in Marrakesh in 1997, which is considered to be one of the main landmarks in this field. Only documents until 2018 were included so that complete annual periods could be compared. In order to avoid duplication, the sample only included original articles [44]. It is worth pointing out that a different search query could give rise to different results. The search was carried out in January 2019. The sample of this study was composed of 713 articles. In addition, a search of articles on "irrigation" was also carried out with the same restrictions in order to analyze the relative importance of sustainability within this general theme. Figure 1 shows an outline of the methodology on which this study was based.

The analyzed variables were the number of articles, their years of publication, all of the authors of the articles, the institutions and countries of all of the authors, the subject areas in which Scopus classifies the studies, the name of the journals in which they were published, and the keywords. After downloading this information, the first task was to eliminate duplications. The names of authors and institutions can be found in different formats. This can lead to errors when counting these records. Therefore, these two variables were analyzed, and the different records were regrouped so that the same author and institution were not counted more than once. Once the information had been refined, different tables and figures were drawn up, and the analysis of the data was conducted. The programs used were Excel (version 2016) and SciMAT (v1.1.04) (University of Granada, Granada, Spain). The tool used to create the network maps was VOSviewer, which is widely used in this type of study [42]. Finally, keyword analysis was used to extract the principal research trends [45]. The terms were regrouped in order to eliminate duplications due to plurals, hyphens, words in upper case letters,

etc. For the grouping of keywords by topics, standardized grouping algorithms were used with the following tools: Vosviewer (Association strength) and SciMAT (network analysis).

**Figure 1.** Outline of the methodological development of this study.

As for the methodology, this work includes some novel aspects compared to previous studies dealing with a similar topic. Firstly, regarding the sample selection of articles to be analyzed, some previous studies made a search based on titles, abstracts, and keywords [41–43,46,47]. In this work, the search was conducted in the fields of title and authors' keywords. Furthermore, before getting the final sample, it was checked that all included articles were related to the actual SI research. Secondly, most works on bibliometric reviews include the analysis of keywords. Nevertheless, this is the first search analysis which aims at detecting SI research trends based on the disciplinary approach of the study and the country where the research was conducted. Finally, the work includes a quantitative review based on a bibliometric analysis, as well as a qualitative one based on the traditional review.
