Layout 1 INTRODUCTION The salinity of the oceans is one of the key parameters for physical oceanography (Dinnat et al., 2019), one of the main variables for monitoring activities and modeling ocean circulation (Klemas, 2011; Falco et al., 2022). Salinity is fundamental in global thermohaline ocean cir- culation, hydrological cycle and climate variability (Boutin et al., 2021). Salinity and temperature control the density of seawater. Density influences the stratification of the ocean, the formation of water masses and finally the general circulation of the ocean itself. Due to its role, sea surface salinity (SSS) has been identified by the Global Climate Observing System (GCOS) as an essential variable for climate studies (Aulicino et al., 2018). De- spite its well-recognized importance, there are few long- term time series measurements of salinity collected in situ, covering the global ocean (Klemas, 2011). Until the early 2000s, most observations and sampling took place through commercial navigation or during oceanographic campaigns. Despite the initial advantage of studying the salinity in the ocean, those methods revealed limitations in terms of spatial and temporal coverage. The acquisition of salinity information has recorded the first major im- provement from the second half of the 2000s onwards, thanks to a network of floating instruments left adrift (Dinnat et al., 2019). However, the most important change occurred with the first satellite observations of the oceans. In recent decades, thanks to the launch of two satellite missions, namely ESA Soil Moisture and Ocean Salinity (SMOS) in 2009 and NASA Aquarius in 2011, it has been possible to monitor the global synoptic of sea surface salinity (SSS) using microwave radiometers (Kerr et al., 2010; Le Vine et al., 2010). A microwave radiometer is a passive sensor capable of measuring the power naturally emitted by a body at a physical temperature above 0 K, proportional to the so-called brightness temperature (Ulaby et al., 1981). For the SSS detection, the two men- tioned satellites use L-band radiometers, from 1,400 to 1,427 GHz, which is reserved for radio astronomy and Earth remote sensing (Pablos and Vall-Llossera, 2014). SMOS and Aquarius operate in the same L-band, but they are different in both the construction of radiometers and their operation (Dinnat et al., 2019; Champagne et al., 2016). The Aquarius mission ended in June 2015 due to a technical malfunction, while the SMOS satellite is still in orbit at the end of 2022. In January 2015, NASA launched a new Soil Moisture Active Passive (SMAP) mission. The SMAP mission aims to measure soil moisture, its ra- diometer-radar instrument is similar to the one used for the Aquarius mission, and it allows SMAP to provide SSS estimates (Bao et al., 2019). SMAP has a spatial resolu- tion of about 40 km with a time repetition of 2-3 days; SMOS spatial resolution is about 45-50 km with a time repetition of 3 days; Aquarius indeed had a resolution of about 100-150 km and a repetition of 7 days (Fournier et al., 2019). Several differences exist in the instrumental approaches, as well as in the retrieval algorithms and in the dielectric constant models (Dinnat et al., 2019), so that the three satellite products present SSS accuracy that gen- erally varies with the sensor and version of the products (Aulicino et al., 2019). Together, they provide an unprece- dented source of salinity information over the global oceans, useful to improve models and compensate for the spatial gaps and the scarcity of in situ observations (Auli- cino et al., 2022). Given these considerations, the purpose of this manu- script is to explore the global scientific literature concern- ing the study of salinity from satellite, to trace its evolution and obtain a systematic review of its temporal developments, the geographical distribution of the main studies and the connections between the different areas of investigation. To this goal, social network analysis and bibliometrics have been combined. The term ‘scientometrics’ refers to the study of the sci- entific literature through measurement and analysis (Pe- rumal and Muthuramalingam, 2021). Bibliometry is a fundamental subfield of scientometrics, and it was first REVIEW Remote sensing of sea surface salinity: A bibliometric analysis Flavia Zanon1, Cinzia Cesarano2, Yuri Cotroneo1, Giannetta Fusco1, Giorgio Budillon1 and Giuseppe Aulicino1,* 1Dipartimento di Scienze e Tecnologie, Università degli Studi di Napoli Parthenope, Napoli; 2Dipartimento di Scienze della Vita e dell’Ambiente, Università Politecnica delle Marche, Ancona, Italy ABSTRACT In recent years, rapid advances in technologies have allowed significant positive changes within the field of satellite observations of the global ocean. This paper reviews the available global scientific literature that focuses on the study of salinity by remote sens- ing, tracking its evolution and trends by combining social network analysis and bibliometrics. Furthermore, the study shows the relationships and co-occurrences between authors, countries and keywords retrieved from the abstracts and citations database pro- vided by Scopus. An analysis of 581 publications has been carried out. The achieved results, which highlight a worldwide increase in scientific interest in this field over the last decade, may constitute a useful tool for a global vision and for a potential improvement in the international efforts employed in the study of salinity from remote sensing. Non -co mmerc ial us e o nly Remote sensing of sea surface salinity: A bibliometric analysis 25 introduced by Pritchard in 1969 (Duan et al., 2020). It deals with the study of the academic literature, through mathematical and statistical methods, with the aim of evaluating global research trends in a specific area of in- vestigation (Geng et al., 2017). Recently, this method has also been applied to the study of remote sensing and its applications in different scientific fields (Hu et al., 2017; Khudzari et al., 2018). However, to date, it has yet to be used to assess the study of salinity by remote sensing. In recent years, the number of publications on the study of SSS by remote sensing has increased remarkably. Therefore, it is important to highlight how the research on this topic is progressing. The final aim of this work is to provide useful information to the scientific community and show potential shortcomings in the research field, as a means of improving future studies and initiatives. The tool used in this work is VOSViewer, a software that al- lows both the realization and visualization of bibliometric maps (Van Eck and Waltman, 2021). The paper is structured as follows: section 2 summa- rizes the criteria related to the bibliographic research car- ried out, as well as the method of bibliometric analysis applied to the data collection; section 3 introduces the re- sults obtained with specific reference to the prevailing au- thors, countries, and keywords within the research; the conclusions are provided in Section 4. MATERIALS AND METHODS Bibliographic research and data collection The bibliometric analysis was performed on a dataset of publications collected through Scopus, Elsevier’s ab- stract and citation database. In our case, the string used was as follows: << “sea surface salinity” AND (satellite OR “remot* sens*”) >>. The Scopus search queried the “Article title”, “Abstract”, “Keywords” database for all the document types published in the English language be- fore December 31st, 2021. The dataset was exported as .csv file on January 31st, 2022, and includes “Citation In- formation”, “Bibliographic Information”, “Abstract”, “Keywords” and “References”. To further refine the data, a careful manual check was carried out to exclude off- topic publications, i.e., studies that were not related to the remotely sensed observation of SSS. The analyses de- scribed in this study were then performed starting from the year of the first publication (i.e., 1976) and focused on remote sensing of SSS indexed in Scopus. Bibliometric network analysis Bibliometric network analysis is a method that is used to understand the research trends within a specific field and based on the results of the scientific literature dataset (Van Eck, 2006). By linking bibliometric data with social network analysis, it is possible to investigate the relation- ships between researchers, countries, organizations, and keywords carried out in the research (e.g., Appolloni et al., 2020; Cesarano et al., 2021). The VOSViewer soft- ware (version 1.6.17) can create cluster maps that are based on the data network, and it also allows us to visu- alize and investigate them (Van Eck and Waltman, 2021). These graphic representations may include different types of outputs (e.g., authors, keywords, journals), and they are realized using citations, bibliographic couplings, co-cita- tions, or co-author relationships (Perumal and Muthura- malingam, 2021). In this study, we examine in depth the relationships between researchers, countries, and key- words. In the cluster maps, the elements’ size depends on the “Total Link Strength” (TLS) (i.e., the strength of one el- ement’s connections with the others), as well as the num- ber of publications and citations. The connections between the different clusters are represented by curved lines where the thickness depends on the “strength of the connections”. Specifically, in the case of co-authorship, the connection strength depends on the number of publi- cations of which two researchers, organizations or coun- tries have “co-authored”; for co-occurrences, instead, it represents the number of times two keywords are paired. Finally, the map resolution determines the number of dis- played clusters. This value was set to 1 for all the analyses presented in this study. As suggested in previous studies (e.g., Khudzari et al., 2018; Appolloni et al., 2020; Ce- sarano et al., 2021), the default settings of VOSviewer were applied for limiting the co-authorship analysis to ar- ticles with a maximum of 25 authors per publication. Among these, we only selected authors and countries with at least five publications. As for keywords, we only processed terms that occurred in at least five of the se- lected publications. RESULTS Bibliometric research and data collection The bibliometric analysis started from the study of the Scopus-indexed publications obtained through the applied research string. The use of the string << “sea surface salin- ity” AND (satellite OR “remot* sens*”) >> proved to be the most inclusive for the literature concerning the study of satellite salinity. Table 1 summarizes the number of available publications and the time period considered. In addition, the number of open-access publications is also indicated; however, this is not a selective criterion in the following analysis. The resulting data were manually ex- amined by all the authors independently to remove off- topic products and to provide a complete and consistent dataset. This process included a title and abstract screen- Non -co mmerc ial us e o nly F. Zanon et al.26 ing, content analysis, and a cross-check of the independ- ent reviews. A total of 581 publications since 1976 were preserved and then included in the bibliometric analysis. Among the excluded publications, the largest percentage deals with SSS ship-based observations or numerical modeling papers in which satellite remote sensing is in- cluded only as an indirect activity (e.g., for providing in- formation about different sea surface parameters, such as temperature or ocean currents) (Table 1). Dataset bibliometric analysis After the pioneer study on remote sensing of SSS at 21-cm wavelength by Thomann (1976), the existence of a few works that date back to the end of 90s (e.g., Le Vine et al., 1998: Njoku et al., 1999) show an early interest in the study of salinity from remote sensing before 2000. Nevertheless, significant concerns from the scientific community seems to be relatively recent. As Figure 1 shows, the time distribution of the 581 publications ana- lyzed reveals a remarkable growth during the last decade. Since 2012, following the launch of the first two satellite missions (i.e., SMOS and Aquarius), and thanks to the first available scientific results, a dramatic increase in the number of publications can be observed (Figure 1). A total of 43 countries have published papers that delve into the study of satellite salinity. The United States (US), a main actor in the Aquarius and SMAP missions, is the leading country in terms of the number of publica- tions indexed in Scopus, namely 263. Then, we can find France (120 publications) and Spain (89), whose interest in the SMOS mission is fundamental. China (86), Italy (36) and the United Kingdom (35) follows (Figure 2a). The geographical distribution by continent (Figure 2b) shows Europe as the leading producer of publications (42%), followed by North America (32%), Asia (20%), South America (4%), Africa and Oceania (1%). Figures 3 and Figure 4 summarize the main institu- tions and researchers involved in these studies. The “CSIC - Consejo Superior de Investigaciones Cientificas”, seems to lead the research in this field, with 140 publications, collecting most of the Spanish efforts in this field of re- search. In the US, the scientific concern is strongly sup- ported by two NASA institutes, the “Jet Propulsion Laboratory” on the West coast, with 69 publications, and the “Goddard Space Flight Center” on the East coast, with 61 publications, together with the “California Institute of Table 1. Summary of the publications results obtained through the search string. Search string Number of publications Time frame Open Access TITLE-ABS-KEY 1057 1976-2021 477 (“sea surface salinity” AND (satellite OR “remot* sens*”) Final refined dataset 581 1976-2021 275 Figure 1. Number of publications indexed in Scopus including the string <<”sea surface salinity” AND (satellite OR “remot* sens*”)>>. Figure 2. Geographical distribution of publications indexed in Scopus (a) per country, including top ten countries, and (b) per continent. Non -co mmerc ial us e o nly Remote sensing of sea surface salinity: A bibliometric analysis 27 Technology”, with 61 publications as well. France, in- stead, shows very different research circumstances. Even though there is strong collaboration and interconnection between groups, research is formally fragmented among different institutes, such as the “Centre National de la Recherche Scientifique” (56), the “Institut de Recherche pour le Developpement” (51), the “Sorbonne Universite” (51) and the “Institut Francais de Recherche pour l’Ex- ploitation de la Mer” (49) (Figure 3). Among the top authors, Prof Jacqueline Boutin emerges as a pioneer and main reference in the subject (Figure 4). She is currently the research director at Sor- bonne Universite/Laboratoire d’Océanographie et du Cli- mat-Expérimentations et Approches Numériques (LOCEAN), Paris, where the exploitation of SSS from the SMOS satellite mission represents one of her main inter- ests. The following authors mostly belong to French and Spanish research groups, constituting an important net- work of collaboration. Thanks to the Scopus database we have investigated other bibliometric characteristics such as the types, scien- tific areas, and sources of indexed publications. Articles account for most publications indexed in Scopus (65%), followed by Conference Papers (31%) Reviews (2%), Books and Book Chapters (1%), and Letters, Notes, Edi- torials or Erratum (1%) (Figure 5). Table 2 lists the top fifteen publication sources for SSS remote sensing stud- ies. “The International Geoscience and Remote Sensing Symposium” (IGARSS) is the first in the ranking with 80 publications (15.2%), followed by the “Journal of Geo- Figure 3. Top ten affiliations associated with publications indexed in Scopus. Table 2. Top fifteen publication sources (number of publications indexed in Scopus). Source Publications Percentage of total publications International Geoscience And Remote Sensing Symposium IGARSS 80 15.2% Journal Of Geophysical Research Oceans 60 11.4% Remote Sensing 54 10.3% IEEE Transactions On Geoscience And Remote Sensing 42 8.0% Remote Sensing Of Environment 32 6.1% Proceedings Of SPIE The International Society For Optical Engineering 28 5.3% Geophysical Research Letters 18 3.4% IEEE Journal Of Selected Topics In Applied Earth Observations And Remote Sensing 18 3.5% International Journal Of Remote Sensing 17 3.2% Journal Of Atmospheric And Oceanic Technology 16 3.0% IEEE Geoscience And Remote Sensing Letters 11 2.1% Ocean Science 7 1.3% Acta Oceanologica Sinica 6 1.1% European Space Agency Special Publication ESA SP 6 1.1% Iop Conference Series Earth And Environmental Science 6 1.1% Non -co mmerc ial us e o nly F. Zanon et al.28 physical Research Oceans” (11.4%), “Remote Sensing” (10.3%), “IEEE Transactions On Geoscience And Remote Sensing” (8%), and “Remote Sensing Of Environment” (6.1%). Figure 6 provides the temporal trend of these top five journals indexed in Scopus, pointing out the remark- able increase in the total number of publications recorded during recent years. Finally, as expected, the subject areas analysis (Table 3) highlights the predominant role of “Earth and Planetary Sciences” (47.5%), followed by “Engineering” (15%) and “Computer Science” (13.8%). Nonetheless, it also shows that the analyzed theme is addressed through a multidis- ciplinary approach, declined in the framework of several different subjects. Bibliometric network analysis The bibliometric network analysis was carried out on the final refined Scopus dataset as previously described in Section 2. Authors The authors, co-authorship analysis included 1278 au- thors. Analyzing only authors with at least five publications, the number lowered to 107 researchers. These can be grouped into eight main clusters. The top fifteen authors are summarized in Table 4, where TLS, links, number of publications and citations are also reported. As expected, Table 4 is generally consistent with the list of authors per number of publications indexed in Scopus presented in Fig- ure 4. From the bibliometric analysis, several clusters are well delineated, with some being interconnected and others maintaining their independence (Figure 7). Indeed, some authors show strong collaborations within their own clus- ters but not with other ones. The main cluster is represented by the red network that groups mostly French scientists belonging to the SMOS community. The yellow cluster is populated by scientists involved in the exploitation of SMOS SSS products too, in this case generally from Spanish research institutes. The blue network also includes several Spanish scientists but pointing out the collaborations that grew up around Prof Jordi Font and Adriano Camps efforts in the devel- opment, preparation and improvement of the SMOS mis- Figure 4. Top fifteen authors publications indexed in Scopus. Figure 5. Publications focused on SSS remote sensing indexed in Scopus by type (in percentage). Figure 6. Publications of the top five scientific journals focused on SSS remote sensing indexed in Scopus per year by source between 1999 and 2021. Table 3. Percentage of publications indexed in Scopus by sub- ject area. Subject area Percentage of total publications Earth and Planetary Sciences 47.5% Engineering 15.0% Computer Science 13.8% Agricultural and Biological Sciences 6.4% Physics and Astronomy 5.5% Environmental Science 3.5% Mathematics 3.2% Materials Science 3.1% Social Sciences 0.8% Others 1.1% Non -co mmerc ial us e o nly Remote sensing of sea surface salinity: A bibliometric analysis 29 sion and its Microwave Imaging Radiometer with Aper- ture Synthesis (MIRAS) instrument. These three clusters are deeply interconnected and highlights the existing in- tense collaboration among the grouped scientist. Two minor clusters (i.e., the cyan and orange ones) are also connected to the main three blocks; they mostly include the ocean salinity researchers belonging to the Chinese scientific community. Conversely, the green and purple large clusters group the US scientists involved in this field of research. Prof Tong Lee can be identified as the main representative of the green network. Together with Severine Fournier, he also represents the main point of interaction between EU and US ocean salinity communities. This cluster is strongly connected to the purple network that includes eminent experts in the development, improvement and op- timization of the instruments onboard the EU and US ocean salinity satellites (e.g., G.S.E. Lagerloef, D.M. Le Vine and E.P. Dinnat), as well as to the small brown clus- ter (Figure 7). Countries The co-authorship countries network includes 50 countries, 21 with at least five publications indexed in Scopus. As stated above (Figure 2b), the role of the EU countries is predominant (42% of publications). The list of the top fifteen countries resembles the information per number of publications reported in Figure 2a. Neverthe- less, even though US authors provided more publications and obtained a higher number of citations, France occu- pies the first place in terms of TLS (Table 5). This result is confirmed by the network map (Figure 8) that repre- Table 4. Summary of the top fifteen authors per total link strength (TLS) in Scopus indexed publications. Author TLS Links Publications Citations Boutin, J. 203 46 50 1632 Gabarró C. 164 38 23 502 Font J. 157 38 34 2474 Reul N. 148 38 38 1388 Martìnez J. 143 34 23 189 Camps A. 134 33 33 1097 Turiel A. 127 33 27 231 Portabella M. 117 31 19 165 Sabia R. 104 40 24 256 Gonzàlez-Gambau V. 101 23 16 67 Corbella I. 97 24 13 381 Lagerloef G.S.E. 97 31 26 1125 Olmedo E. 93 17 21 165 Martìn-Neira M. 92 28 18 642 Lee T. 90 37 28 641 Figure 7. Co-authorship network map of authors based on total link strength. Colors refer to different clusters. The bigger the circle size the greater total link strength the author has. The closer the circles are the more often the authors are found in the same publications. Figure 8. Co-authorship network map of countries based on total link strength. Colors refer to different clusters. Circle size indi- cates a countries publication frequency. Circle proximity indicates increased occurrence of two countries being found in the same publication. Non -co mmerc ial us e o nly F. Zanon et al.30 sents the way in which countries and clusters are con- nected. From the bibliometric analysis, in fact, five clusters emerge. The remoteness of the connections outlines the scarce collaboration between some countries and a stronger dialogue between others. Crucial is the role of France (blue cluster) and the US (yellow cluster), as we can see from their central position on the map. Although these countries also develop collaborations, the limited number and remoteness of links with other countries, es- pecially for US, suggest that they tend to favor independ- ent national research activities and solid exclusive collaborations within their cluster. Conversely, the rele- vant cluster led by Spain (in red) presents strong intercon- nections between several countries (i.e., UK, Italy, Norway, Denmark, Germany and the Netherlands) and a general equilibrium among the partners. Another cluster is represented by the green network led by China that brings together several countries (i.e., Canada, Australia and Japan) that cannot be ordered according to typically geographical or linguistic criteria, but that share common thematic interests in the framework of scientific research concerning the study of salinity by remote sensing. Fi- nally, Russian Federation represents a self-standing unit with only few links with external countries (i.e., France and US) (Figure 8). Keywords The analysis of the co-occurrence of keywords in- cludes 2899 words, 218 of which with at least five key- word occurrences. From the network analysis five clusters emerge. The term “sea surface salinity” is the most recur- ring keyword, followed by “remote sensing”, “sea sur- face”, “salinity”, “oceanography” and “SMOS” (Table 6). Figure 9 highlights the relevance and the relationships of the keywords. The circles size demonstrates how fre- quently a term appears as keyword in the different publi- cations. Node proximity indicates how the research topics are strongly related to each other, i.e., the frequency of two terms occurring together in either the abstract, title or keyword listing of the publications. The network map also reflects how the matter is investigated. The five clusters are well-defined, and they identify the specific aspects of existing research activities. On the other hand, they are strongly interconnected and partially overlapping. Table 5. Top fifteen countries per total link strength (TLS) in Scopus indexed publications. Country TLS Links Publications Citations France 157 17 122 4369 United States 145 18 267 4709 Spain 111 13 89 3317 Italy 78 8 36 1037 Netherlands 78 10 33 2519 United Kingdom 69 12 35 676 Germany 32 9 14 232 China 23 10 92 548 Argentina 22 2 21 580 India 22 7 32 335 Norway 19 10 8 71 Australia 16 9 10 102 Denmark 13 7 5 49 Benin 9 3 7 108 Brazil 9 6 7 65 Figure 9. Co-occurrence network map of keywords based on total link strength. Colors refer to different clusters, circle size indicates keyword frequency. Circle proximity indicates the frequency of two keywords being found in the same publication. The size of the keyword label is proportional to its total link strength. Non -co mmerc ial us e o nly Remote sensing of sea surface salinity: A bibliometric analysis 31 The main cluster is the red one that has a crucial role; the predominant keyword is “sea surface salinity” around which several connections between this cluster and the oth- ers occur. It groups the words related to the general impor- tance of SSS studies, the water cycle and the climate. Among others, the cluster includes words like “air-sea in- teraction” and “seasonal variability”, “global ocean” and “climatology”, “enso” and “el nino”, “precipitation” and “evaporation”, “vertical mixing” and “stratification, “upper ocean” and “mixed layer”. The second cluster, represented in green, is more fo- cused on the oceanographic aspects; here the dominant words are “oceanography”, “sea surface temperature”, “satellite data”, “in situ measurement”, and “satellite”. Many of these keywords are located in the very center of the map but they are not visible in Figure 9 because covered by stronger items (in terms of TLS) belonging to other clus- ters. Several terms linked to the water quality and the pri- mary productivity are also present here (e.g., “ocean color”, “phytoplankton”, “chl-a”, “cdom”, “river discharge”, “bio- chemistry”, “ecosystems”, “modis”), as well as items re- lated to the satellite data processing and analysis (e.g., “dataset”, “mean square error”, “linear regression”, “least square algorithm”, “estimation method”, “image tech- nique”, “accuracy”, neural network”, “machine learning”). The third cluster is represented by the blue network that groups the terms referring to satellite studies, e.g., “remote sensing”, “microwave”, “radiometer”, “brightness temper- ature”, “l-band”, “interferometry”, “antennas”, “sensitivity analysis”. As mentioned above, these three clusters are deeply in- terconnected and they all present a high number of links to- wards the fourth network, the yellow cluster, whose items are largely shaded by red, blue and green terms. This cluster gathers the words related to the main satellite missions such as “SMAP”, “SMOS”, “active/passive”, “image resolu- tion”, “radio interference”, “quality control”, “interpola- tion”, “rmse”, “uncertainty analysis”. The fifth grouping (in purple) seems to be oriented to the Aquarius mission, as proved by words like “nasa”, “space optics”, “orbits”, “space flight” and, of course, “Aquarius”. As expected, this cluster is strongly related to the yellow one that includes the other two main SSS satel- lite missions (i.e., SMOS and SMAP). DISCUSSION AND CONCLUSIONS In this study, we analyze the scientific literature con- cerning the study of salinity from remote sensing. Global satellite observations led to a better understanding of the influence of salinity on ocean circulation, biogeochem- istry and climate variability, as well as better monitoring of the water cycle (Reul et al., 2020). The first publica- tions indexed in Scopus date back to the late 70s. How- ever, only since 2012 the increase in the number of published articles has shown significant growth in this rel- atively new field of research. Literature analysis usually helps to define strategic de- cisions about which research areas to support and which areas need better management. For these reasons biblio- metrics has assumed considerable relevance in recent years becoming an important tool for science (Reuters, 2008). In this work we specifically explore authors, coun- tries, and keywords with the purpose of providing a global view, highlighting the strengths of this research and pos- sible improvements for the future. The existence of well- defined research groups such as the French, Spanish and US ones emerged. Even though these clusters appear to be strictly interconnected, the analysis of the co-author- ship countries network shows that several countries (e.g., US and France) have a preference for independent na- tional research activities and for exclusive collaborations within their clusters. This suggests that there is a wide possibility to further strengthen links between different countries by encouraging greater cooperation for scien- tific progress in the monitoring of SSS from satellites. The analysis of the co-occurrence of keywords groups the main terms representative of the SSS satellite research in five well-defined clusters that identify the specific as- pects of existing research activities. These clusters are highly consistent in being organized around main topics (i.e., the general importance of SSS studies in the climatic context, the oceanographic aspects, the satellite capabili- ties) and the main missions (i.e., SMOS, SMAP, Aquarius). Moreover, they are strongly interconnected due to the ele- vate number of links crossing the boundary of a single group. This confirms that all the research topics are strongly Table 6. Top fifteen keywords per occurrence in Scopus indexed publications. Keyword TLS Occurrence Sea surface salinity 4444 465 Sea surface 2791 248 Remote Sensing 2787 282 Salinity 2443 242 Oceanography 2248 195 Soil moisture 1875 158 SMOS 1771 165 Radiometer 1701 153 Microwave 1504 145 Sea surface temperature 1441 136 Aquarius 1406 138 Satellite 1152 97 Satellite data 1013 96 Brightness temperature 892 74 Algorithm 782 70 Non -co mmerc ial us e o nly F. Zanon et al.32 related to each other and that the engineering aspects are evolving in synergy with the scientific results. As suggested by Vinogradova et al., (2019), future satellite SSS missions will benefit from this synergistic approach and from con- solidating strong partnerships based on the common pur- pose of advancing salinity studies and applications for societal benefit. Nevertheless, from a deeper analysis of the selected ar- ticles some problems regarding the ability to study SSS by remote sensing emerge. One of the main issues concerns the validation of the existing products, probably due to the difficulties in the comparison between in situ and satellite observations (Boutin et al., 2016). Satellite data are inte- grated over a large surface, and they generally provide daily, weekly, or monthly information. Conversely, in situ data are punctual in space and time. In addition, satellite salinity is measured at the surface level while in situ salinity is generally measured at a few meters’ depth. Recently, Thouvenin-Masson et al. (2022) assessed that this sampling difference predicts a different sensitivity of the instruments to small-scale variability. This problem is particularly evi- dent in regions with high SSS variability. Unfortunately, to- gether with the intrinsic spatio-temporal variability of available L-band observations, this issue limits the potential use of satellite SSS retrievals in other fields of research (e.g., polar and coastal oceanography, air-sea interaction, marine biology and ecology, sea pollution) that would be affected by accuracy and gaps of the available SSS prod- ucts. Also, the analyzed literature shows that these issues still limit the incorporation of SSS data collected through satellites into environmental models for their initialization and assimilation, so that this field of research is still at an early stage (Vinogradova et al., 2019). In recent years statistical analyses, machine learning, image reconstruction and technological innovation have tried to address these issues and led to great progress in the SSS satellite research (e.g., Olmedo et al., 2018; Buon- giorno-Nardelli, 2020; Boutin et al., 2021; Guimbard et al., 2021; Li et al., 2022; Jang et al., 2021; Bao et al., 2021). SSS satellite data are now frequently assimilated in opera- tional numerical weather prediction systems, helping to im- prove the forecasting of the global hydrological cycle. However, many improvements are still necessary to im- prove this area of investigation, which is particularly impor- tant for the future study and understanding of oceanographic and climate dynamics, as well as of the biological and eco- logical systems. To improve remote sensing of SSS in cold regions, for example, additional studies should focus on P- band frequencies where, differently from L-band, the sen- sitivity to salinity is nearly invariant with water temperature (Vinogradova et al., 2019). Furthermore, to improve the consistency of low microwave frequency observations across all oceans and temperatures, future research should also focus on seawater dielectric constant and atmospheric attenuation (Dinnat et al., 2019). These efforts will also re- duce dependency on empirical corrections. To these goals, additional ground truth SSS information is certainly desir- able. An interesting summary of useful recommendations for salinity remote sensing for the next decade has been pro- vided by Vinogradova et al. (2019) (see Table 1). They in- clude: ensuring the continuity of space-based SSS measurements; improving accuracy, resolution, and cover- age of satellite SSS observing systems, especially in polar oceans; advancing the integration of the observed SSS into global ocean observing network and modeling/assimilation; improving the understanding of SSS data uncertainties and the effects of sampling differences from in situ measure- ments; developing innovative, cost-effective solutions for future satellite missions and exploring multi-frequency in- strument concept to enable simultaneous measurements of various ocean parameters; pursuing international collabora- tions in terms of technology, calibration/validation frame- work, and cost-sharing. Corresponding author: Giuseppe Aulicino, Centro Direzionale Isola C4, 80143, Napoli, Italy. Tel. +390815476586. E-mail: giuseppe.aulicino@uniparthenope.it. Authors’ contributions: All the authors made a substantive intel- lectual contribution, performed part of the experiments. All the au- thors have read and approved the final version of the manuscript and agreed to be accountable for all aspects of the work. Conflict of interest: The authors declare no potential conflict of in- terest. Funding: This work was partly funded by the Italian Ministry of Uni- versity and Research (MUR) under the Antarctica National Research Programme (PNRA) entitled ”SWIMMING - Sea ice-wave interac- tion monitoring for marginal ice navigation”, PNRA18_00298. Availability of data and materials: All data generated or analyzed during this study are included in this published article. Key words: sea surface salinity; remote sensing of the ocean; bib- liometric network analysis. Received: 9 September 2022. Accepted: 3 November 2022. Publisher’s note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher. ©Copyright: the Author(s), 2022 Licensee PAGEPress, Italy Advances in Oceanography and Limnology, 2022; 13:10862 DOI: 10.4081/aiol.2022.10862 This work is licensed under a Creative Commons Attribution- NonCommercial 4.0 International License (CC BY-NC 4.0). Non -co mmerc ial us e o nly Remote sensing of sea surface salinity: A bibliometric analysis 33 REFERENCES Appolloni L, Buonocore E, Russo GF, Franzese PP. (2020). The use of remote sensing for monitoring Posidonia oceanica and Marine Protected Areas: A systemic review. Ecological Ques- tions 31:7-17. Aulicino G, Cotroneo Y, Olmedo E, Cesarano C, Fusco G, Budil- lon G. (2019). In Situ and Satellite Sea Surface Salinity in the Algerian Basin Observed through ABACUS Glider Measure- ments and BEC SMOS Regional Products. Remote Sensing, 11:1361. Aulicino G, Cotroneo Y, De Ruggiero P, Buono A, Corcione V, Nunziata F et al. (2022). Remote Sensing Applications in Satellite Oceanography. Springer Series in Measurement Sci- ence and Technology. Springer Cham, Midtown Manhattan, USA. p. 181-209. Aulicino G, Cotroneo Y, Ansorge I, Van Den Berg M, Cesarano C, Rivas MB, et al. (2018). Sea surface salinity and temperature in the southern Atlantic Ocean from South African icebreakers, 2010-2017. Earth System Science Data. 10:1227- 1236. Bao S, Wang H, Zhang R, Yan H, Chen J. (2019). Comparison of Satellite-Derived Sea Surface Salinity Products from SMOS, Aquarius, and SMAP. Journal of Geophysical Research: Oceans. 124:1932-1944. Bao S, Zhang R, Wang H, Yan H, Chen J, Wang Y. (2021). Cor- rection of Satellite Sea Surface Salinity Products Using En- semble Learning Method. IEEE Access. 99:1-1. Boutin J, Chao Y, Asher WE, Delcroix T, Drucker R, Drushka K, et al. (2016). Satellite and in Situ Salinity: Understanding Near-Surface Stratification and Subfootprint Variability. Bul- letin of the American Meteorological Society. 97:1391-1407. Boutin J, Reul N, Koehler J, Martin A, Catany R, Guimbard S, et al. (2021). Satellite-based sea surface salinity designed for ocean and climate studies. Journal of Geophysical Research: Oceans. 126:e2021JC017676. Boutin J, Reul N, Kohler J, Martin ACH, Catany R, Guimbard S, et al. (2021). Satellite-based Time-Series of Sea Surface Salin- ity designed for Ocean and Climate Studies. Journal of Geo- physical Research Ocean. 126. Buongiorno Nardelli B. (2020). A Deep Learning Network to Re- trieve Ocean Hydrographic Profiles from Combined Satellite and In Situ Measurements. Remote Sensing. 12:3151. Champagne C, Rowlandson T, Berg A, et al. (2016). Satellite sur- face soil moisture from SMOS and Aquarius: Assessment for applications in agricultural landscapes. International Journal of Applied Earth Observation and Geoinformation. 143-154. Dinnat EP, Le Vine DM, Boutin J, Meissner T, Lagerloef G. (2019). Remote Sensing of Sea Surface Salinity: Comparison of Satellite and In Situ Observations and Impact of Retrieval Parameters. Remote sensing. 11: 750. Duan P, Wang Y, Yin P. Remote Sensing Applications in Moni- toring of Protected Areas: A Bibliometric Analysis. Remote sensing. 12:772. Cesarano C, Aulicino G, Cerrano C, Ponti M, Puce S. (2021). Sci- entific knowledge on marine beach litter: A bibliometric analysis. Marine Pollution Bulletin. 173. Cui X, Guo X, Wang Y, Wang X, Zhu W, Shie J, et al. Application of remote sensing to water environmental processes under a changing climate. Journal of Hydrology. 574:892–902. Falco P, Castagno P, Cotroneo Y, Aulicino G, Budillon G, De Rug- giero P, et al. (2022). Measurements for Oceanography. Springer Series in Measurement Science and Technology. Springer Cham, Midtown Manhattan, USA. p. 51-81. Fournier S, Lee T, Tang W, Steele M, Olmedo E. (2019). Evalua- tion and Intercomparison of SMOS, Aquarius, and SMAP Sea Surface Salinity Products in the Arctic Ocean. Remote Sens- ing. 11:3043. Geng Y, Chen W, Liu Z, Chui ASF, Han W, Liu Z, et al. (2017). A bibliometric review: Energy consumption and greenhouse gas emission in the residential sector. Journal of Cleaner Pro- duction. 159:301–316. Guimbard S, Reul N, Sabia R, Herlédan S, Khoury Hanna ZE, Piollé JF, et al. (2021). The Salinity Pilot-Mission Exploitation Platform (Pi-MEP): A Hub for Validation and Exploitation of Satellite Sea Surface Salinity Data. Remote Sensing. 13:4600. Hu K, Qi K, Guan Q, Wu C, Yu J, Qing Y, et al. (2017). A Scien- tometric Visualization Analysis for Night-Time Light Remote Sensing Research from 1991 to 2016. Remote Sensing. 9:802. Jang E, Kim YJ, Im J, Park YG. (2021). Improvement of SMAP sea surface salinity in river-dominated oceans using machine learning approaches. GIScience & Remote Sens- ing. 58:138-160. Kerr Y, Waldteufel P, Wigneron JP, Delwart S, Cabot F, Boutin J, et al. (2010). The SMOS mission: A new tool for monitoring key elements of the global water cycle. Proceedings of the IEEE. 98:666–687. Klemas V. (2011). Remote Sensing of Sea Surface Salinity: An Overview with Case Studies. Journal of Coastal Research. 27:830–838. Khudzari JM, Kurian J, Tartakovsky B, Vijaya Raghavan GS. (2018). Bibliometric analysis of global research trends on mi- crobial fuel cells using Scopus database. Biochemical Engi- neering Journal. 51-60. Le Vine DM, Kao M, Garvine RW, Sanders T. (1998). Remote Sensing of Ocean Salinity: Results from the Delaware Coastal Current Experiment. Journal of Atmospheric and Oceanic Technology. 15:1478-1484. Le Vine D, Lagerloef G, Torrusio S. (2010). Aquarius and remote sensing of sea surface salinity from space. Proceedings of the IEEE. 98:688–703. Li Y, Yin X, Zhou W, Lin M, Liu H, Li Y. (2022). Performance Simulation of the Payload IMR and MICAP Onboard the Chi- nese Ocean Salinity Satellite. IEEE Transactions on Geo- science and Remote Sensing. 60:1-16. Njoku EG, Rahmat-Samii Y, Sercel J, Wilson WJ, Moghaddam M. (1999). Evaluation of an inflatable antenna concept for microwave sensing of soil moisture and ocean salinity. IEEE Transactions on Geoscience and Remote Sensing. 371:63-78. Olmedo E, Gabarró C, González-Gambau V, Martínez J, Ballabr- era-Poy J, Turiel A, et al. (2018). Seven Years of SMOS Sea Surface Salinity at High Latitudes: Variability in Arctic and Sub-Arctic Regions. Remote Sensing. 10:1772. Pablos M, Vall-Llossera M. (2014). SMOS and Aquarius Ra- diometers: Inter-Comparison Over Selected Targets. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. 7:9. Perumal K, Muthuramalingam S. (2021). Global sources, abun- dance, size, and distribution of microplastics in marine sedi- ments - A critical review. Estuarine, Coastal and Shelf Science. 264. Non -co mmerc ial us e o nly F. Zanon et al.34 Reul N, Grodsky SA, Arias M, et al. (2020). Sea surface salinity estimates from spaceborne L-band radiometers: An overview of the first decade of observation (2010-2019). Remote Sens- ing of Environment. 242. Reuters T. (2008). Whitepaper using bibliometrics: a guide to evaluating research performance with citation data. Available at: http://openscience.ens.fr/MARIE_FARGE/CONFER- ENCES/2014_12_02_BIBLIOMETRIE_ET_EVALUA- TION_DE_LA_RECHERCHE_ABDU_PARIS/InCites_Tho mson-Reuters.pdf Thomann GC. (1976). Experimental Results of the Remote Sens- ing of Sea-Surface Salinity at 21-cm Wavelength. IEEE Transactions on Geoscience Electronics. 14:198-214. Thouvenin-Masson C, Boutin J, Vergely JL, et al. (2022). Satellite and in situ sampling mismatches: consequences for the esti- mate of satellite sea surface salinity uncertainties. Remote Sensing. 14. Ulaby FT, Moore RK, Fung K. (1981). Fundamentals and radiom- etry. Microwave Remote Sensing Active and Passive. Artech House Publishers, London, UK. 456 pp. Van Eck N, Waltman L, Van den Berg J, Kaymak U. (2006). Vi- sualizing the computational intelligence field. IEEE Compu- tational Intelligence Magazine. 1:6–10. Van Eck NJ, Waltman L. (2021). VOSviewer Manual 1.6.17. Vinogradova N, Lee T, Boutin J, Drushka K, Fournier S, Sabia R, et al. (2019). Satellite Salinity Observing System: Recent Dis- coveries and the Way Forward. Frontiers in Marine Science. 6:243. Non -co mmerc ial us e o nly