Dispersion and Disparity: Bibliometric and Visualized Analysis of Research on Climate Change Science Communication

Research on climate change science communication began in the 1980s and is showing continued vitality and a wider interest at present. In order to track the development of global research on the communication of climate change hot topics and frontier progress since the 21st century, methods such as bibliometrics and co-word network analysis were used to analyze the publication of research papers in this field, and a total of 1175 valid papers published in 2000–2021 in the WOS core database were counted. Different dimensions such as temporal trend, spatial distribution, and author collaboration network were analyzed. The results show that, (1) climate change communication research has become a relatively independent research field and has entered a rapid development stage, and this field still has a broad research prospect in the new understanding of climate change and new international context. (2) At present, research in this field is still dominated by developed countries, but developing countries are actively building their unique climate communication discourse. (3) Public understanding and media information presentation have been hot topics in climate communication research in recent years. In the context of changing international situations and the development of global epidemics and new climate policies, changes in national actions will likely lead to new research topics and dialogues. Research shows that climate change science communication research is increasingly showing a trend of decentralization and differentiation.


Introduction
Since Swedish researcher and Nobel laureate Svante Arrhenius published his report on human-caused global warming in 1896 [1], climate change has entered the scope of scientific research. Climate change was defined as a purely environmental issue in the early perceptions of scholars worldwide, and before the 1970s, meteorological scientists believed that climate characteristics could be described in units of 30-year average climate conditions. It was not until the 1990s that the dynamic and nonindependent nature of the climate change system gradually gained academic recognition, and the abruptness of climate change reduced the time unit for characterizing climate to a decade or even less [2,3].
The phenomenon of climate communication has emerged almost simultaneously with the issue of climate change, and climate communication has played an important role in the response to climate change. Thus, in early climate communication practices, scientific findings, extreme weather events, political summits, and regular reports issued by the Intergovernmental Panel on Climate Change (IPCC) were given more coverage [4]. As the issue of climate change has received increasing social attention, research on climate communication has been on the rise. Climate communication researchers generally believe

Data Sources and Collection Procedure
The literature source of this paper is the Web of Science core collection database, from which we collected data on April 5, 2022. To ensure that first-hand data are comprehensive, have a high degree of interpretation, and as many samples as possible are obtained, topic research rather than advanced research was used. The literature was searched with the formula of "TS = science communication of climate change". The first search conducted produced 1933 articles, and then the literature was filtered according to the time distribution of the number of articles, and the time span was set from 1 January 2000 to 31 December 2021. Considering the quality of literature and cross-sectional comparison factors, the sample literature was subjected to data cleaning work, excluding review papers, conference proceedings papers, book reviews, editorial materials and unrelated disciplinary literature, and a total of 1559 literature were retrieved, and 1175 valid literature were obtained after further screening based on research topics and literature content.
In the analysis of collaborative networks and general statistics, we set the node type as author, institution, and country, respectively, and used the cosine algorithm as the calculation of linkage strength. In the analysis of research hotspots and frontiers, keywords were selected as node type, noun terms were extracted from the original keywords of authors and supplementary keywords of the dataset to name the clusters, and the loglikelihood ratio (LLR) was used for the extraction of cluster labels. The time slice in all operation steps was set to 1 year, and the threshold (k-value) was set to 25. The specific research route of this study is shown in Figure 1.

Research Methodology and Tools
Bibliography can be traced back to 1917 [14] when Hume first introduced the concept of "Statistical Bibliography" in 1923, and by extending the object of study from academic journals to all available literature, bibliographers developed bibliography into bibliometrics [15,16].
In order to achieve a bibliometric analysis of the literature, in addition to statistical analysis of the time and country related to the literature using descriptive statistics, keyword co-occurrence analysis and author collaboration network analysis were used to analyze the collaboration of hot topics and publications published in this research area and to track important research areas and developments.
This study uses CiteSpace analysis software in the field of bibliometrics. Based on a path-finding network algorithm (pathfinder), as well as co-citation analysis theory (cocitation), CiteSpace is capable of identifying the research progress, new trends and new developments in a certain field in scientific development and presenting them in a visual

Research Methodology and Tools
Bibliography can be traced back to 1917 [14] when Hume first introduced the concept of "Statistical Bibliography" in 1923, and by extending the object of study from academic journals to all available literature, bibliographers developed bibliography into bibliometrics [15,16].
In order to achieve a bibliometric analysis of the literature, in addition to statistical analysis of the time and country related to the literature using descriptive statistics, keyword co-occurrence analysis and author collaboration network analysis were used to analyze the collaboration of hot topics and publications published in this research area and to track important research areas and developments.
The number of annual publications can reflect the level of development of a research field to determine its research development [22]. Based on the historical literature included in the Web of Science (Figure 2), it can be found that the topic of climate change science communication did not receive much attention from academics until 2007. Since 2008, research on this area has entered a phase of rapid growth, reaching a peak in 2021. In terms of the time scale of the number of publications, climate change communication research can be divided into three stages: the preliminary stage, the stable growth stage, and the rapid development stage. In the preliminary phase (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007), only 39 articles appeared in eight years, which represents 3.3% of all the literature. During the five years of the steady growth phase (2008-2013), 194 articles were published, almost five times the total number of articles in the first phase, accounting for 16.5% of the total literature. This indicates that researchers during this period began to show interest in climate change communication by rapidly publishing research results and occupying priority in scientific discovery. In the third period (2014-2021), the number of publications occupies 80.2% of the total literature, with an average of 117.8 publications per year. This indicates that climate change communication research has formed an independent research direction and become a more popular research area.
The increase in research literature from 2008 onward is mainly attributed to the following two factors: (1) the advancement of natural science, which has revealed the impact of human activities on climate change and the deterioration of the human living environment; sensing the environmental crisis of human existence has led social scientists to explore ways to promote environment-friendly behavior [23]. (2) The social construction of science. More and more countries and nonprofit organizations have become aware of the relationship between climate change and human life and have promoted negotiation and collaboration between countries, and the high level of international and public concern has triggered researchers to pay attention to the communication of climate change [8]. The Copenhagen Conference held in 2009 is a typical example. After this conference, research outputs related to climate change communication show a significant increase and the research in this field began to enter a new era of rapid development, raising the attention of governments, the public, enterprises, and other diversified social actors so that climate change became the frontier of research. The increase in research literature from 2008 onward is mainly attributed to the following two factors: (1) the advancement of natural science, which has revealed the impact of human activities on climate change and the deterioration of the human living environment; sensing the environmental crisis of human existence has led social scientists to explore ways to promote environment-friendly behavior [23]. (2) The social construction of science. More and more countries and nonprofit organizations have become aware of the relationship between climate change and human life and have promoted negotiation and collaboration between countries, and the high level of international and public concern has triggered researchers to pay attention to the communication of climate change [8]. The Copenhagen Conference held in 2009 is a typical example. After this conference, research outputs related to climate change communication show a significant increase and the research in this field began to enter a new era of rapid development, raising the attention of governments, the public, enterprises, and other diversified social actors so that climate change became the frontier of research.
According to Price's law, many indicators in science grow according to an exponential law. In other words, the growth rate of various scientific indicators is proportional to the existing values of the indicators. The exponential curve estimation analysis using SPSS 25.0 software yielded the exponential function of the curve as: The results showed that R 2 = 0.791, p < 0.001, which is consistent with Price's law, indicating that climate change communication research is still in a rapid development phase at this time.

Spatial Distribution of Studies
Statistics on the countries in which the literature is published can reflect the importance of the country and its influence in a particular field of research [24]. When different authors, institutions and countries appear in the same paper, CiteSpace assumes that a collaborative relationship was created in the writing of the paper. Therefore, all countries where the authors were located are included in the statistics. If multiple authors of a paper were from the same country, this country was counted only once. Figure 3 plots the distribution of literature output by country in the scientific research of climate change communication, and research in this domain has been mainly concentrated in Europe and North America, which echoes the earlier qualitative studies [25]. Among them, the U.S. is not only the first country to study climate communication issues, but also the region with, by far, the highest number of publications. Although the UK, Germany, Australia and Canada are among the top 5 countries with some research accumulation with the U.S., there is still a large gap between the four countries and the U.S. The number of studies in Asia, South America and Africa are generally low, and scholars from countries other than the United States have focused on this area and conducted research relatively late, mainly between 2008 and 2010.
Among the top 20 countries to which the authors of the papers belonged (Table 1), all of them were developed countries except China and South Africa. This indicates that there is a correlation between the level of economic development and climate change communication research, developed countries are more concerned with scientific communication research on climate change. On one hand, developed countries generally have stronger academic research capacity and academic innovation than developing countries, and are also more involved in the theoretical exploration of climate communication at an earlier stage. On the other hand, for developing countries, economic development and poverty eradication are the primary goals of these countries in the industrialization process, so in the practice of scientific research, research related to climate change communication was conducted later than in developed countries. However, as the response to climate change becomes a global governance activity, developing countries are also actively involved in this action and related research, and their research focused on building a framework for climate communication discourse in developing countries, for example, China as the representative of developing countries, hoping to fight for more space for survival and development for developing countries in international climate negotiations and to assume common but differentiated responsibilities [26][27][28][29].  Among the top 20 countries to which the authors of the papers belonged (Table 1), all of them were developed countries except China and South Africa. This indicates that there is a correlation between the level of economic development and climate change communication research, developed countries are more concerned with scientific communication research on climate change. On one hand, developed countries generally have stronger academic research capacity and academic innovation than developing countries, and are also more involved in the theoretical exploration of climate communication at an earlier stage. On the other hand, for developing countries, economic development and poverty eradication are the primary goals of these countries in the industrialization process, so in the practice of scientific research, research related to climate change communication was conducted later than in developed countries. However, as the response to climate change becomes a global governance activity, developing countries are also actively involved in this action and related research, and their research focused on building a framework for climate communication discourse in developing countries, for example, China as the representative of developing countries, hoping to fight for more space for survival and development for developing countries in international climate negotiations and to assume common but differentiated responsibilities [26][27][28][29].

Analysis of Core Authors
Authors are the foundation of the research field and the source of innovation, and core authors represent a cluster of researchers who have contributed significantly to the improvement of a certain research field. Table 2 shows the top 20 scholars in the field of climate communication research in terms of the number of publications. The well-known bibliometrician Price proposed that authors with m or more publications are outstanding scientists, i.e., core authors, calculated by the formula [30]: In the above formula, n max is the number of articles published by the highest-producing author. Anthony Leiserowitz is the most prolific author in the area of climate change communication, with a cumulative total of nine publications, from which the value of m is calculated to be 2.247. After rounding off, all authors with a number of publications greater than or equal to 3 can be considered core authors in the field of climate communication, and there are 16 of them.
From the analysis of the statistical results, it is clear that the total number of publications by core authors of climate communication research between 2000 and 2021 is 67, accounting for only 5.7% of the total number of publications, which is far from the requirement that half of the papers in this field are written by core authors as stated by Price. Therefore, there is currently no large and stable core group of authors in the worldwide climate change science communication research community. In total, 14 scholars have single articles with 20 or more citations, and Kahan's study, published in 2012 and cited in 38 papers to date, suggests that public disinterest and disagreement about climate change stems from a unique conflict of interest that includes individual and collective interests [31]. Another equally highly cited article comes from Moser, who in 2016 reviewed important advances and emerging trends in climate communication and received extensive attention [6]. Unlike selective literature analysis, Cook and his team have received the same peer attention by analyzing the evolution of the scientific consensus on anthropogenic global warming (AGW) in the scientific literature through a bibliometric approach [32].

Collaborative Network of Authors
The analysis and visual presentation of the network structure among authors can clearly demonstrate the cooperative relationships among scholars and institutions, and provide a new perspective for the evaluation of academic impact in this field. From Figure 4, we can see that there are 526 authors, and the number of collaborations is 505. The collaboration density (D) is presented by the ratio of the actual number of relationships (m) of a node to the theoretical maximum number of relationships in a network with n nodes, which is 0.0037. The calculation formula is [33]: The analysis and visual presentation of the network structure among authors can clearly demonstrate the cooperative relationships among scholars and institutions, and provide a new perspective for the evaluation of academic impact in this field. From Figure  4, we can see that there are 526 authors, and the number of collaborations is 505. The collaboration density (D) is presented by the ratio of the actual number of relationships (m) of a node to the theoretical maximum number of relationships in a network with n nodes, which is 0.0037. The calculation formula is [33]: In the case of low overall collaboration density but more collaborative relationships, it indicates that a few articles contain a large number of the author-author collaborations, which also indicates that most author collaborations in the research of climate change communication are dominated by small-scale collaborations and individual studies. Thus, the collaboration among author groups is expected to be further strengthened.
Through our analysis, we found that a stable, long-term collaborative network has formed among the core authors. For example, Leiserowitz, Gustafson and Goldberg experimentally test how perceptions of natural gas vary depending on what it is called; they also analyzed how targeted advertising shifts Republican views on climate change [34]. Leiserowitz and Maibach also investigated public perceptions of the human health risks of climate communication in the United States, Canada, and Malta [35].
Although the worldwide collaboration of researchers working on the topic of climate change science communication is fragmented, the field of climate communication has produced several stable collaborative networks (Figure 4), namely Tarla Peterson and Andrea Feldpausch-Parker, Franck Courchamp and Celine Bellard, and Michael Brueggemann In the case of low overall collaboration density but more collaborative relationships, it indicates that a few articles contain a large number of the author-author collaborations, which also indicates that most author collaborations in the research of climate change communication are dominated by small-scale collaborations and individual studies. Thus, the collaboration among author groups is expected to be further strengthened.
Through our analysis, we found that a stable, long-term collaborative network has formed among the core authors. For example, Leiserowitz, Gustafson and Goldberg experimentally test how perceptions of natural gas vary depending on what it is called; they also analyzed how targeted advertising shifts Republican views on climate change [34]. Leiserowitz and Maibach also investigated public perceptions of the human health risks of climate communication in the United States, Canada, and Malta [35].
Although the worldwide collaboration of researchers working on the topic of climate change science communication is fragmented, the field of climate communication has produced several stable collaborative networks (Figure 4), namely Tarla Peterson and Andrea Feldpausch-Parker, Franck Courchamp and Celine Bellard, and Michael Brueggemann and Stefanie Walter. The node size is proportional to the number of articles posted by the authors, and the connections between nodes represent how closely the authors collaborate. Michael and Stefanie have co-authored a total of six articles, four of which are about group perceptions of climate change and scientists' media-use behavior in climate change. Jaspal has established a collaboration with two other researchers, Nerlich and Koteyko, and they examined the rhetorical aspects of social debates about climate change in the Daily Mail's reader comments [36]. In addition, they collaborated to analyze remarks posted on the British tabloid internet site at two different points in time earlier and later than the East Anglia controversy [37]. Sebestyen, Czvetko and Abonyi highlight the applicability of data and models in complex climate change research issues [38].

Institutions and Cooperation Network
Collaborative network mapping was used to analyze the main contributing institutions and the relationship between them. The findings indicate that the relatively influential institutions in the climate change communication research field are concentrated in the UK ( Table 3). The U.S. and other developed countries, with George Mason University, Cornell University, Oxford University, and Cardiff University, started to focus on climate communication issues before 2010, having accumulated rich research results in this research area. It is noteworthy that the Chinese Academy of Sciences, which first began to emerge in the research results on climate change science communication in 2015, has accumulated 17 papers so far, ranking sixth, with rapid growth in results. The above mentioned institutions, together with the University of British Columbia, the University of Michigan, and Yale University, have an extremely important position in the research field of climate change science communication, and cooperate more closely with other research institutions, assuming an important linkage role. The lines emitted from a node indicate the collaboration between nodes, and larger nodes and more lines mean a closer research collaboration. By setting the node type to institution, we calculated the number of emissions and cooperation networks of research institutions ( Figure 5). The analysis found that 955 connecting lines were formed between 438 institutions, with a cooperation density of 0.1. This indicates that there is closer cooperation between different scientific research institutions. George Mason University has accumulated some theoretical results in differentiated audience research, including elites, cooperative extension professionals, students, and adolescents [39][40][41][42]. Cornell University tends to look at ways to more effectively promote the practice of climate change science communication by developing a climate change information system or new approach for evaluating climate change communication. What is more, they examined the link between coffee consumption and climate change communication and found that the coffee consumption experience can be regarded as a practical tool for public participation in climate change [43,44]. The Chinese Academy of Sciences uses bibliometrics to study research trends in environmental communication and considers the lack of international cooperation as a major shortcoming in the field of environmental communication research [45].

Keyword Co-Occurrence Analysis
The frequency of co-occurrence refers to the number of times a node appears in the network, and the frequency of co-occurrence directly reflects the research hotspot [46]. The size of the tree ring reflects the frequency of keywords. The co-occurrence analysis of keywords ( Figure 6) revealed that "climate change", "science", "communication ", "science communication", "knowledge" and "policy" are high-frequency terms from 2000 to 2021. This reveals that knowledge about climate change, knowledge diffusion and climate change-related policies were the hot topics of research during this period. From Table 4, we can see that "climate change" was the earliest and the most frequent term, with a frequency of 645. Science and communication were the second and third keywords, respectively, and both appeared for the first time in 2005. There are 8 keywords with frequencies greater than 100, and 15 words with frequencies above 60. In addition to climate change and science communication always being the core keywords in this field, how to reduce social risks and uncertainties, and public attitudes and behaviors have also become popular topics of interest for researchers.

Keyword Co-Occurrence Analysis
The frequency of co-occurrence refers to the number of times a node appears in the network, and the frequency of co-occurrence directly reflects the research hotspot [46]. The size of the tree ring reflects the frequency of keywords. The co-occurrence analysis of keywords ( Figure 6) revealed that "climate change", "science", "communication ", "science communication", "knowledge" and "policy" are high-frequency terms from 2000 to 2021. This reveals that knowledge about climate change, knowledge diffusion and climate changerelated policies were the hot topics of research during this period. From Table 4, we can see that "climate change" was the earliest and the most frequent term, with a frequency of 645. Science and communication were the second and third keywords, respectively, and both appeared for the first time in 2005. There are 8 keywords with frequencies greater than 100, and 15 words with frequencies above 60. In addition to climate change and science communication always being the core keywords in this field, how to reduce social risks and uncertainties, and public attitudes and behaviors have also become popular topics of interest for researchers.     1  climate change  645  2000  2  science  424  2005  3  communication  213  2005  4  knowledge  157  2004   5  science  communication  145  2009   6  policy  118  2005  7  impact  104  2003  8  perception  103  2003  9  risk  99  2003  10  management  97  2002  11  information  89  2005  12  uncertainty  72  2003  13  adaptation  72  2003  14  media  66  2005  15 attitude 61 2006

Analysis of Main Topics
Keyword clustering can reflect the research clusters formed in a subject area by clustering closely related keywords into one category. The data showed that the keyword clustering network density for climate change propagation in keyword clustering and timeline analysis was 0.021, Silhouette values characterizing keyword homogeneity were 0.6777, and Modularity values were greater than 0.3, which indicated high homogeneity of the mapping network and reasonable clustering results.
Literature in the same cluster is placed in the same color block, and the more literature in a cluster means that the more important the cluster area is. The closely related keywords were grouped into 10 major clusters (Figure 7), and the clusters were divided into three main categories to provide a clearer picture of the main research themes in the field of climate change communication.
clustering network density for climate change propagation in keyword clustering and timeline analysis was 0.021, Silhouette values characterizing keyword homogeneity were 0.6777, and Modularity values were greater than 0.3, which indicated high homogeneity of the mapping network and reasonable clustering results. Literature in the same cluster is placed in the same color block, and the more literature in a cluster means that the more important the cluster area is. The closely related keywords were grouped into 10 major clusters (Figure 7), and the clusters were divided into three main categories to provide a clearer picture of the main research themes in the field of climate change communication. (1) Various topics based on climate change in the natural sciences (cluster#5, clus-ter#6, and cluster#9). Due to the impact of climate change on all aspects of human life, scholars have generated a variety of research topics around climate change in their respective research fields, but in general, they hope to promote the understanding of climate change in the scientific community and the public through their respective research. For example, in the field of science education, Bell Wayne and others from Washington College use real environmental data lessons to improve students' understanding of basic science concepts. In the area of the Internet, Salam, A. has attempted to build a framework for the Internet of Things, hoping to enhance the understanding of regional climate through the use of technological solutions [47,48]. (1) Various topics based on climate change in the natural sciences (cluster#5, cluster#6, and cluster#9). Due to the impact of climate change on all aspects of human life, scholars have generated a variety of research topics around climate change in their respective research fields, but in general, they hope to promote the understanding of climate change in the scientific community and the public through their respective research. For example, in the field of science education, Bell Wayne and others from Washington College use real environmental data lessons to improve students' understanding of basic science concepts. In the area of the Internet, Salam, A. has attempted to build a framework for the Internet of Things, hoping to enhance the understanding of regional climate through the use of technological solutions [47,48].
(2) Climate Change Science Communication (cluster#1, cluster#2, cluster#3, and clus-ter#10). This is central to the field of climate communication, with environmental communication, social science, and rhetoric as keywords in this category, and researchers have worked to identify how messages are communicated and the role of media [49,50]. They also investigate how social science methods and practices can raise awareness of climate change policy, economics and culture among student groups, and engage them in thinking about the topic of climate change and science communication [51,52].
(3) Public Policy (cluster#4, cluster#7, and cluster#8). Because climate change poses a multifaceted and complex threat to human health and well-being, researchers want to alert and examine how to communicate this threat and health risks to the public [53].For instance, exploring possible pathways for changing dietary habits and health behaviors to reduce carbon emissions through programmed behaviors to slow the rate of climate degradation [54,55]. Others compared policies to address climate change in different countries around the globe and explored and identified key government departments and decisions that have had a significant impact on the climate change process, examining the role of climate change communication in the political discourse of sustainable development policies [56,57].

Analysis of Research Frontiers
CiteSpace can detect the sudden emergence or fading of literature keywords in the research field within a period of time through the bibliometric statistical calculation method, and this detection method can visually present the frontier knowledge hotspots and their evolutionary trends in the field. By identifying and tracking the emergent keywords, we can better understand the evolution of research hotspots over time and judge the development trend at the same time, from which we can predict the scientific problems that may emerge in the future which need to be solved urgently, and then provide the information as a reference for the future direction of efforts. By using the burst detection function of CiteSpace, the obtained literature was statistically analyzed, and the burst graphs of a total of 15 keywords were obtained by sorting them by the burst year and excluding the keywords that grew or faded rapidly within a short period of time ( Figure 8). The red part represents the keywords that are emergent in a certain year or years, and the blue represents the keywords that are non-emergent in that time region. alert and examine how to communicate this threat and health risks to the public [53].F instance, exploring possible pathways for changing dietary habits and health behaviors reduce carbon emissions through programmed behaviors to slow the rate of climate d radation [54,55]. Others compared policies to address climate change in different countr around the globe and explored and identified key government departments and decisio that have had a significant impact on the climate change process, examining the role climate change communication in the political discourse of sustainable development p icies [56,57].

Analysis of Research Frontiers
CiteSpace can detect the sudden emergence or fading of literature keywords in research field within a period of time through the bibliometric statistical calculat method, and this detection method can visually present the frontier knowledge hotsp and their evolutionary trends in the field. By identifying and tracking the emergent k words, we can better understand the evolution of research hotspots over time and jud the development trend at the same time, from which we can predict the scientific pr lems that may emerge in the future which need to be solved urgently, and then prov the information as a reference for the future direction of efforts. By using the burst det tion function of CiteSpace, the obtained literature was statistically analyzed, and the bu graphs of a total of 15 keywords were obtained by sorting them by the burst year a excluding the keywords that grew or faded rapidly within a short period of time (Fig  8). The red part represents the keywords that are emergent in a certain year or years, a the blue represents the keywords that are non-emergent in that time region. The keyword statistics show that the top 5 keywords in climate change communication research between 2000 and 2021 are "representation", "United States", "news", "view" and "media representation". The main research hotspots in climate communication can be divided into three phases according to the duration of the heat of the emergent words.
The first phase, from 2003-2013, focused on responses relating to natural climate change rather than climate communication, how news coverage of climate change influenced public perceptions of uncertainty in scientific findings [58], and the impact of global environmental assessments on developed and developing countries [59,60].
In the second phase (from 2008-2016), climate communication practices and strategies received more attention, and a variety of communication tools such as narratives and rhetoric of games and climate communication public works were included in the study [61,62]. Beyond the public, researchers began to focus on broader climate communication practices, such as how early risk communication and consensus should be reached between scientists, planners, and policymakers [63,64].
In the third phase, 2011-2018, people's perceptions and climate-resilient behavior became the focus of research, and American scholar Oregon proposed a dialogic com-munication model of communication that would enable participants to gain a deeper understanding of knowledge, beliefs, perceptions, values, and barriers relating to climate communication [65].
Media presentation has emergent words at different stages, which we believe is largely attributable to the iterations of media technology and related to the important role that the media has assumed in public-facing climate communication. Newspapers, television, news, and social media became the arenas for studying climate communication practices [66][67][68][69]. In general, the early areas of climate communication research were more focused, and after 2004, the areas of research have become more dispersed and diverse.

Discussion
This paper takes the research literature on climate change science communication published in the WOS core database from 2000-2021 as the object of study, uses CiteSpace to systematically analyze the development history of international climate communication research, and summarizes the research findings through data analysis as follows: (1) For the general trend of research, the field of climate communication research has shown rapid development, with increasing research heat and expansion of influence. The relevant literature was relatively small before 2007, with no more than five annual publications, and it entered a phase of rapid growth from 2008, with more than 100 research articles published for the first time in 2018. The change in the volume of publications is closely related to the fact that human actions continue to influence and accelerate the rate of change in the Earth's environment and climate, and that natural events (such as the increasing frequency of extreme weather) and social events (international climate conferences and national climate frameworks, etc.) have attracted the attention of scholars, all of which have driven further research on climate communication.
(2) In terms of the spatial distribution of literature output, researchers in developed countries are the main population engaged in climate communication research and have accumulated rich academic results in this field, while developing countries, represented by China and South Africa, are actively involved in research on this topic.
(3) Research on climate change in general encompasses natural science understanding of climate change, policy research on climate change, and research on the communication of climate change in science.In recent years, public perception, adaptive management and media presentation have been hot topics of research in the field of climate communication.
(4) Last but not least, research in this field is dominated by individual research and collaboration within small groups. Cooperation and collision between different disciplines, institutions and even countries are likely to produce more quality research.

Conclusions
The overall study shows that climate communication is showing strong momentum in the present and future..The number of publications in the first four months of 2021 already surpassed the number of publications for the whole year of 2020 and the average for the third period (2014-2021). We believe that this research area is currently showing great potential and scope.
The extent to which developing countries influence the global climate change process is not proportional to their ability to resist global climate change, and since there are huge differences between developing and developed countries in multiple latitudes such as economic development, energy structure, climate characteristics, and the level of scientific literacy of citizens, under the framework of the United Nations Framework Convention on Climate Change, researchers in developing countries should pay more attention and effort to the theory and practice research in climate change communication.
While the global pandemic and the war may have opened up interesting analogs for this field, with the emergence of new climate policies, new national agreements, and new communication technologies, policy implications, effective case studies, and the impact of electronic and virtual media on public perceptions, attitudes, and behaviors are likely to be new research topics.
Funding: This research received no external funding.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The experiment data used to support the findings of this study are included in the article.

Conflicts of Interest:
The authors declare no conflict of interest.