1. Introduction
Remote Sensing Open Access Journal (RS OAJ) is an international peer-reviewed scholarly open access journal established in the year 2009 and published every month by MDPIAG, SWITZERLAND. It publishes regular research papers, technical notes or letters, review articles, and communications. Today, the journal is very well recognized in remote sensing science and technology and other spatial sciences such as the geographic information systems (GIS), global positioning systems (GPS), and global navigation satellite system (GNSS). A wide array of remote sensing subjects are covered that include passive remote sensing from sensors such as multispectral and hyperspectral sensors [
1,
2], microwave radiometer (Yavuz, Teixeira, 2009) [
3], thermal radiometer (Gao, Kustas & Anderson, 2012) [
4], etc. Active remote sensing from sensors such as Lidar (Colgan, Baldeck & Féret et al., 2012) [
5], Radar (Joshi, Baumann & Ehammer et al., 2016) [
6], Sonar (Hasan, Ierodiaconou & Monk, 2012) [
7], Scatterometer (Zwieback, Paulik & Wagner, 2015) [
8], Altimeter (Bosch, Dettmering & Schwatke, 2014) [
9]. The Remote Sensing Open Access Journal (RS OAJ) of the MDPI is indexed by Science Citation Index Expanded, Ei Compendex, SCOPUS and some other famous databases. In the 2017 Journal Citation Reports (JCR) released by Web of Science, it had an impact factor (IF) of 3.406 and a CiteScore (Elsevier) of 4.03.
This year (2018) is the 10th anniversary of RS OAJ and a special issue is organized to celebrate this. In this milestone, conducting a general bibliometric review about RS OAJ is particularly pertinent and valuable. It is common to publish special issues (Meyer, Winer, 2014) [
10] when a journal is holding a significant anniversary. In particular, a bibliometric overview of the journal is noteworthy because it provides some historical results and the retrospective evaluation of the journal is presented for us (Schwert, 1993) [
11]. Research on bibliometric of a journal has developed for a long time (Heck, Bremser, 1986) [
12]. Many scholars have done bibliometric analysis of journals during the anniversary celebrations. Such as Shugan (2006) [
13] developed a bibliometric overview of the journal Marketing Science. Van Fleet (2006) [
14] wrote an article about The Journal of Management’s First 30 Years published in the Journal of Management. Merigó (2015) [
15] of the papers published Journal of Business Research. Cancino (2017) [
16] presented an overview of the Computers & Industrial Engineering. Merigó (2017) [
17] developed a bibliometric analysis of the thirty years of International Journal of Intelligent Systems. Valenzuela (2017) [
18] presented an overview of Journal of Business & Industrial Marketing. Tang (2018) [
19] wrote an article about Ten Years of Sustainability (2009 to 2018): A Bibliometric Overview.
Bibliometrics provides us with a tool that can be easily extended from the micro to the macro level. Bibliometric indicators are increasingly used as research performance evaluation tools. These indicators are based on bibliographic databases designed primarily for information retrieval purposes. Such as co-citation, journal impact factor, total cites, eigenfactor score, normalized eigenfactor, CiteScore, h-index. We can understand the characteristics of journals macroscopically through these indicators.
The RS OAJ published by MDPI has not been systematically reviewed by bibliometric method previously. Therefore, in this article, we conduct a comprehensive bibliometric profile of RS OAJ that will help answer questions like:
- (1)
What are the dynamics and trends of RS OAJ publications over last 10-years?
- (2)
What are the journal impact factor, total cites, eigenfactor score, normalized eigenfactor, CiteScore of RS OAJ and the publications speed of various remote sensing journals?
- (3)
What is the h-index of RS OAJ, and how are the h-classic publications distributed?
- (4)
What are the major institutions and countries (or territories) according to number of publications, and the cooperation patterns among them?
- (5)
What are the main research themes?
- (6)
What are the citation impact of co-occurrences keywords?
- (7)
What is the intellectual structure analysis about RS OAJ? and
- (8)
What is the knowledge commutation analysis about RS OAJ?
In each of the above, a comparison is made with the similar factors of other leading remote sensing journals of the world.
This article is expected to achieve several goals. First, it is expected to help readers to get a quick, intuitive, and profound overview of RS OAJ, and help relevant scientists/readers decide whether or not to contribute articles to the RS OAJ. Second, exploring the research status of RS OAJ by bibliometric analysis, some meaningful information will be provided to improve the visibility of RS OAJ. Third, a comparison with other leading remote sensing journals will highlight strengths and limitations of RS OAJ. Fourth, a comprehensive review of this nature will help take stock, get critical feedback from the scientific community, understand what has been done right so far and help focus on advancing the journal to next level. A comprehensive understanding of the journal is achieved by analyzing factors such as the number of citations, most cited papers, influential authors, document types, impact factor (IF), the publication years, the most productive institutions and countries (territories). Fifth, In August 2005, Jorge Hirsch (2005) [
20] proposed a research performance indicator called h-index to measure the scientific performance of scholars. It is defined as follows, “
A scientist has index h
if h
of his or her N
p papers have at least h
citations each and the other (N
p − h
) of papers have ≤h
citations each”. (Hirsch, 2005) [
20]. So, an h-index of 25 means that a scientist has 25 papers that are each cited atleast 25 times. This new indicator has attracted great interest in the field of informetrics, scientometrics and bibliometrics. Butler & McAllister (2011) [
21] had confirmed the applicability of h-index to social science researchers. Costas & Bordons (2007) [
22] confirmed the h-index, which includes the total number of publications and the citation of these publications, has recently been proposed as an objective criterion for academic productivity. The advantage of the h-index is that it gives a robust estimate of the wide impact of scientists’ cumulative research contributions (Hirsch, 2005) [
20]. The authors of this manuscript suggest to take a more nuanced assessment of h-index. This calls to consider h-index based on the authorship ranking; that is whether the one is the first, second, third, and so on to nth author. The synthesis will develop visualization tools that are employed to exhibit the development characteristics of RS OAJ and compare the same with other leading remote sensing journals. In addition, the h-index can not only be applied to assessment achievement of a single researcher, but also can be applied to academic journals (Bornmann, 2005) [
23].
2. Materials and Methods
The data for this article was retrieved from the Web of Science database after comparing it with other databases as it contains panoramic information of RS OAJ. We used the journal title = “Remote Sensing” to search for publications. There were 5573 publications in total from 2009 until the date of search (2 August 2018). In addition, there were 15 publications missing from the database, but found in the RS OAJ journal homepage (which included 8 editorials, 5 new book reviews, and 2 articles). We processed these 15 publications according to the Web of Science format, and a total of 5588 publications were obtained. RS OAJ contains 7 types of publications: research articles (5373, 96.15%), review articles (103, 1.84%), editorials (57, 1.02%), corrections/addendum (44, 0.79%), letters (4, 0.07%), book reviews (5, 0.14%) and biographies (2, 0.04%).
Bibliometric analysis is an effective way to study and test a knowledge field (Braun, Schubert, 2003) [
24] and it also can avoid subjective judgment (Garfield, 1972) [
25]. A large number of bibliometric methods are used to evaluate research performance. Because this is a quantitative study, providing many indicators to assess the literature (Broadus, 1987) [
26]. This article used a wide range of bibliometric indicators, including the dynamics and trends of publications, h-index, h-classic publications, co-authorship countries (territories) and institutions, citation impact of co-occurrences keywords, co-occurrences author keywords and other related indicators (Alonso, Cabrerizo, Herrera-Viedma & Herrera, 2009; Franceschini, Maisano, 2010; Hirsch, 2005) [
20,
27,
28]. Since Henry Small (1980) [
29] introduced the concept of co-citation for the first time and used the node link network to visualize the co-citation relationship of 10 famous particle physics papers, a large number of studies have applied the visualization of the co-citation relationship. In a series of subsequent co-citation studies (Small 1981 and 2006) [
30,
31], Boyack (2014) [
32] and White (2014) [
33] studied the principle of co-citation analysis and its application in the process of scientific development and determined the dynamic intellectual structure of science as a whole or in a specific field. The scholars expanded the analysis unit from the paper to the author, resulting in the author co-citation analysis (ACA) (Nerur, 2010) [
34]. Through a lot of self-reflection research on co-citation research, two main types of co-citation analysis, that is, DCA and ACA, can be found to visualize the whole or specific domain intelligent structure (Chen, Ibekwe-Sanjuan & Hou, 2010) [
35].
We made a comparison with the 20 other leading remote sensing journals (
Table 1) taking four factors: journal impact factor, total cites, eigenfactor score, and normalized eigenfactor. We have considered other equivalent journals which overwhelmingly publish remote sensing data, methods, and science.
Journal Impact Factor means number of times all the articles that are published in the last two years (e.g., 2015, 2016) in a journal are cited this year (e.g., 2017) by any journal in the JCR database divided by the total number of articles published [
36,
37].
where “k” is a year, “Nk − 1 + Nk − 2” is the number of papers published by the journal in the previous two years, “nk − 1” and “nk − 2” represent the number of citations of the journal in year “k”.
Specifically, the journal impact factor of RS OAJ in 2017 can be calculated as the following formula:
Total cites means the total number of times that a journal has been cited by all journals included in the database in the Journal Citation Reports (JCR) year [
38].
Eigenfactor score calculation is based on the number of times articles from the journal published in the past five years have been cited in the JCR year, but it also considers which journals have contributed these citations so that highly cited journals will influence the network more than lesser cited journals [
39,
40]. Remote Sensing Open Access Journal (RS OAJ) with eigenfactor score of 0.0342 is next only to Remote Sensing of Environment (0.0529), and IEEE Transaction of Geoscience and Remote Sensing (0.0434), showing that RS OAJ has excellent record of its articles cited in other high ranked journals.
The normalized eigenfactor Score is the eigenfactor score normalized, by rescaling the total number of journals in the JCR each year, so that the average journal has a score of 1. Journals can then be compared and influence measured by their score relative to 1. For example, Remote Sensing Journal of MDPI which has a normalized eigenfactor of 3.99, means that the journal is 3.99 times more influential as the average journal in the Journal Citation Report (JCR) [
41].
Another measure of the impact of the journals is depicted by the CiteScore. CiteScore is a new journal evaluation index published by Elsevier publishers in 2016. CiteScore calculates the average number of citations received in a calendar year by all items published in that journal in the preceding three years. Papers published in journals for 3 consecutive years are cited times in the fourth year without excluding any type of articles. It increases the citation period by one year compared with the impact factors. (
https://journalmetrics.scopus.com) [
42].
We also analysis the most productive countries (territories) and institutions. Because we can see the distribution characteristics of high-yielding countries and institutions, we can find out which countries and institutions are investing most of their efforts in remote sensing. Our research method is to obtain the most productive countries (territories) and institutions by downloading RS OAJ data from the Web of science database and sorting them in descending order. In addition, through the VOSviewer, we can analyze the amount of cooperation between countries. Further explore which countries have closer cooperation. In a nutshell, we use bibliometric and cartography to explore the bibliometric characteristics of the RS OAJ in this article. And we use h-index and h-classic articles to identify highly cited articles in RS OAJ. We also used co-citation to analyze intellectual structure. In addition, we use VOSviewer with the visual intelligence structure tools (Van Eck et al., 2010) [
43]. This is because VOSviewer makes it more intuitive to display panoramic information about RS OAJ (Bonilla, Merigó & Torres-Abad, 2015; Ding, Rousseau & Wolfram, 2014) [
44,
45].
4. Comparison with Two Best Remote Sensing Journals
The two best, well known, remote sensing journals are Remote Sensing of Environment or RSE (published: 1969–current) and the IEEE Transactions on Geoscience and Remote Sensing (IEEE TGRS) (published: 1980–current) (Previous Title: IEEE Transactions on Geoscience Electronics; published: 1963–1979) It is interesting to compare Remote Sensing Open Access Journal of MDPI (RS OAJ) (published: 2009–current) with these two best remote sensing journals (
Table 15). Year 2016 was chosen due to availability of complete data for all three journals obtained from Web of Science. RS OAJ published about 2.3 times number of articles when compared with RSE and about1.8 times that of IEEE TGRS. Overall, RS OAJ has highest number of cites (6248), followed by RSE (5854), and IEEE TGRS (5740). However, the average cites per item is just the reverse: RSE (13.07), IEEE TGRS (10.02), and RS OAJ (6.03) (
Table 15). However, RS OAJ has 28 papers with 21 or greater citations compared to 62 papers and 83 papers with citations of 21 or higher in IEEE TGRS and RSE respectively (
Table 9). Also, RS OAJ has 136 papers with 11–20 citations compared to 92 papers and 114 papers with citations of 10–21 in IEEE TGRS and RSE respectively (
Table 15). On the low end, about 7% of the papers published in RS OAJ do not have any citations. In comparison those percentages for IEEE TGRS and RSE were about 5.0% and 2.5% respectively (
Table 15). Also, about 22% of the papers published in RS OAJ have just 1–2 citations. In comparison those percentages for IEEE TGRS and RSE were about 19% and 8% respectively (
Table 15). So, it is clear that if RS OAJ can eliminate publishing a good percentage of likely low cited articles in comping years, its impact factor can swiftly rise. However, RS OAJ philosophy is also to publish all good papers that have value, and technically sound but may not be cited frequently due to various reasons.
We also made detailed statistics on the citations of three journals from 2010 to 2017. In
Table 16, the median citation number for RSE for 2010 is 42 (95%: 186), for RS OAJ the median is 15 (95%: 92), for IEEE TGRS the median is 23 (95%: 112). RSE had in 2010 a total number of 244 publications. We found the following 95% percentile: 186 citations and the following median (50% percentile): 42 citations. IEEE TGRS had in 2010 a total number of 383 publications. We calculated the following 95% percentile: 112 citations and the following median: 23 citations. RS OAJ had in 2010 a total number of 143 publications. We calculated the following 95% percentile: 92 citations and the following median: 15 citations. We also calculated the number of articles with citations less than 3 that found the percentage of citations less than 3 in each year from 2010 to 2013 was about 5, and the percentage increased in 2014–2017. This is a good indication that there are too many papers that don’t get (or little) attention. The possible reason is that it takes some time to quote, and the other reason is that these articles are of “poor” quality. In order to become Top 1 journal, RS OAJ would need to significantly reduce the number of “poor” papers in the future.
5. Discussion
In 2018, RS OAJ will celebrate its 10th anniversary. Motivated by this event, this article proposes a bibliometric and visualization methods to analyze the main trends of RS OAJ from 2009 to 2018 (years in which it is published). On the basis of the indicators put forward by Cancino (2017) [
16], Tang (2018) [
19], Voner (2016) [
86] and Merigó (2018) [
87], our article analyzes a wide array of different type of bibliometric indicators, including dynamics and trends of publications, journal impact factor, total cites, eigenfactor score, normalized eigenfactor, CiteScore, h-index, h-classic publications, most productive countries (territories) and institutions, co-authorship collaboration about countries (territories), research themes, citation impact of co-occurrences keywords, intellectual structure and knowledge commutation, which are reveal the bibliometric characteristics of the journal [
88].
Our article shows that the number of publications published in RS OAJ has been increasing. The highest number of published publications was in 2017 (1336 publications), 298 publications more than in 2016 (1038 publications). By analyzing the trend of citations per publication, the annual citation trend of publications is stable from 2009 to 2012 with an average citation rate of 26 per publication. However, we can see that from 2013 to 2018, the average number of citations per publication has annually decreased. The possible reason is the number of publications published in recent years (2013–2018), as it takes time to cite them. At the same time, editors and reviewers should continue to strictly review the procedures so that improve quality of RS OAJ. Based on 2016 data, about 7% of the papers published in RS OAJ do not have any citations. In comparison those percentages for IEEE TGRS and RSE were 5.0% and 2.5% respectively. Also, about 22% of the papers published in RS OAJ have just 1–2 citations. In comparison percentages of 1–2 citations for articles published in IEEE TGRS and RSE were about 19% and 8% respectively. This shows that if RS OAJ can reduce a certain articles types of articles that are not likely to be cited, its impact factor can increase dramatically. However, this is a tradeoff between publishing some good articles that may lack novelty and hence less likely to be quoted, versus publishing them for certain value they offer (e.g., a drought study conducted in another region using a well-known method). Impact factor were proposed by Garfield and Sher (1963) [
89], and was used to rank and evaluate journals (Garfield, 1996) [
90]. We can see that RS OAJ has an impressive journal impact factor of 3.4060 for the year 2017. During 2017, a total of 13,600 times the articles published in RS OAJ are cited in the journals included in the Journal Citation Reports (JCR). The eigenfactor score of 0.0342 and normalized eigenfactor score of 3.9902. For the year 2017, the CiteScore for RS OAJ is 4.03.
Through analyzing h-index, RS OAJ has an h-index of 67 from 2009–2018. Considering h-classic publications, the most cited article title is Global Data Sets of Vegetation Leaf Area Index (LAI)3g and Fraction of Photosynthetically Active Radiation (FPAR)3g Derived from Global Inventory Modeling and Mapping Studies (GIMMS) Normalized Difference Vegetation Index (NDVI3g) for the Period 1981 to 2011. China and the United States have cooperated most frequently, probably the number of publications published are largest in both countries. The most productive country is China (2012 publications) but there are few publications from China are highly cited as shown in
Table 3. Thus, Chinese scholars should pay more attention to improve the quality of publications rather than the quantity in future research. By analyzing most productive institutions, we find that the first four prolific institutions are located in China. Through further analysis, compared to other journals that belongs to remote sensing. Such as, Remote Sensing of Environment (United States, 3804 publications; France, 778 publications), IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (China, 2031 publications; United States, 741 publications), Journal of Applied Remote Sensing (China, 1462 publications; United States, 561 publications) and ISPRS Journal of Photogrammetry and Remote Sensing (China, 428 publications; United States, 375 publications). This indicate China and the United States are the two countries that have published the most publications on remote sensing. In the high-frequency author keywords analysis, the major co-occurrences author keywords are remote sensing, MODIS, Landsat, LiDAR, NDVI, classification, hyperspectral, soil moisture, SAR and validation. For further study, we cluster analysis the author’s keywords and get 5 research themes: Multi-spectral and hyperspectral remote sensing, LiDAR scanning and forestry remote sensing monitoring, MODIS and LAI data applications, Remote sensing applications and Synthetic Aperture Radar (SAR).
Through author keywords citation impact analysis, the most influential keyword is Unmanned Aerial Vehicle (UAV), this indicate the publications about Unmanned Aerial Vehicle (UAV) attract more citations than other keywords in RS OAJ. Followed, forestry, Normalized Difference Vegetation Index (NDVI), terrestrial laser scanning, airborne laser scanning, forestry inventory, urban heat island, monitoring, agriculture, and laser scanning. By analyzing the intellectual structure of RS OAJ, we identify the main reference publications and find that the themes are about Random Forests, MODIS vegetation indices, image analysis, remotely sensed data, Soil-Adjusted Vegetation Index (SAVI) and SAR interferometry etc. RS OAJ ranks first in cited journals and third in citing, this indicates that RS OAJ has the internal knowledge flow. This is called the self-citation of the journal and the self-citation rate of RS OAJ is 24.10%. Brown considered it is a common phenomenon for a journal to cite itself [
91].
Limitations
Some of the limitations of this article deserve to be noted. First, due to the bibliometric method, we mainly use frequencies plotted in charts to show the status of RS OAJ, This is because the frequency is the most commonly index in the bibliometric methods. Charts intuitively display the statistical information (time, quantity, etc.), and it is a very good method to vividly show information. However, this may lead us to neglect some valuable information. Such as centrality, degree centrality, and effective scale. So such indicators are worth pursuing in future research. Second, we analyzed the indicators (including dynamics and trends of publications, journal impact factor, total cites, eigenfactor score, normalized eigenfactor, CiteScore, h-index, h-classic publications, most productive countries (territories) and institutions, co-authorship collaboration about countries (territories), research themes, citation impact of co-occurrences keywords, intellectual structure and knowledge commutation). However, we did not analyze the correlation between the indicators, which may be worth looking at in future.
6. Conclusions
This paper presents a comprehensive bibliometric profile of the Remote Sensing Open Access Journal (RS OAJ) based on its publication years (2009–2018). During these 10 years, there has been an exponential growth in the number of articles published, going from around 100 articles in 2009 and 2010 to 1336 articles in 2017 and reaching about the same in 2018. During 2009–2018, there were 129 countries and 3826 institutions that published 5588 articles. The leading nations contributing articles, based on 2009–2018 data, were (based on ranking): China, United States, Germany, Italy, France, Spain, Canada, England, Australia, Netherlands, Japan, Switzerland and Austria. The leading institutions, also for the same period and listed based on ranking, were: Chinese Academy of Sciences, Wuhan University, University of Chinese Academy of Sciences, Beijing Normal University, The university of Maryland, National Aeronautics and Space Administration, National Oceanic and Atmospheric Administration, China University of Geosciences, United States Geological Survey, German Aerospace Centre, University of Twente, and California Institute of Technology.
The h-index of RS OAJ based on data from 2010 to 2017 (data of full years of publication) was 67. For the same period, there were two remote sensing journals with higher h-index: Remote Sensing of Environment with (h-index = 112), and IEEE Transactions on Geoscience and Remote Sensing (h-index = 101). For 2017, the latest year for which impact data are available, the RS OAJ had journal impact factor of 3.4060, and a CiteScore of 4.03. Further, RS OAJ had eigenfactor score of 0.0342 for the year 2017, which was next only to Remote Sensing of Environment or RSE (0.0529), and IEEE Transaction of Geoscience and Remote Sensing (0.0434) amongst the best remote sensing journals. This shows that RS OAJ has excellent record of its articles being cited in other high ranked journals. The normalized eigenfactor score of 3.99 for RS OAJ during the year 2017, means that the journal is 3.99 times more influential as the average journal in the Journal Citation Report (JCR). There were a total of 12,327 journals in JCR during 2017. Also, for the year 2017, the CiteScore for RS OAJ was 4.03. CiteScore calculates the average number of citations received in a calendar year by all items published in that journal in the preceding three years.
Other comparisons were also made to gauge impact based on recent years’ data. The 2016 data showed, RS OAJ published about 2.3 times number of articles when compared with RSE and about 1.8 times that of IEEE TGRS, the two leading international remote sensing journals. Overall, RS OAJ has highest number of cites (6248), followed by RSE (5854), and IEEE TGRS (5740). However, the average cites per article is just the reverse: RSE (13.07), IEEE TGRS (10.02), and RS OAJ (6.03). The trends are about the same for 2017. During 2009–2018, there were 814 keyword themes that occurred 5 or more times in the published articles. The paper established these topics as well as the most frequently occurring topics that were published in RS OAJ. The paper also shows the knowledge flow into and from RS OAJ to other journals and proposes a nuanced h-index (nh-index) to measure productivity and intellectual contribution of authors by considering h-index based on whether the one is first, second, third, or nth author.