The economics of Forest Carbon Sequestration: A Bibliometric Analysis

Carbon sequestration in forests has increasingly captured the attention of scientists as a strategy for climate change mitigation and environmental sustainability. In this era of huge carbon emission, being a low-carbon and cost-effective technology, the economic analysis of forest carbon sequestration holds higher importance for the successful implementation and intended outcomes. This study elucidates a scientometric view of the research structure and thematic evolution of economic studies on forest carbon sequestration based on 1,439 articles over the time slice 2001-2021. The bibliographic data has been retrieved from the Dimensions database which accommodates a large coverage of research publications and also provides easy access to essential scholarly data and information. Vosviewer and Biblioshiny software tools have opted for visualization and evaluation purposes of bibliometric data. This study employs various measures of bibliometric analysis like co-authorship, bibliographic coupling, citation and keyword analysis to nd out the principal articles, authors, journals, most frequent keywords and highest publishing countries and institutions in this eld and the results show that the number of publications has escalated substantially in the last ve years, Popp A, 2017 (305 citations) and André P C Faaij (11 documents) are the most cited article and the most productive author, respectively, Bradford’s law calculates 21 core journals out of total 503 journals among which Forest Policy and Economics is on the top, and the most productive country and institution are the USA and University of Florida, respectively. The study also investigates key publishing subject categories and the number of publications covered under each Sustainable Development Goals. The overall outcome of this bibliometric study confers an in-depth understanding of the various dimensions of economic analysis on forest carbon sequestration, its development pattern in the last 20 years and also provides emerging themes for future references.


Introduction
In this era, unsustainable rapid development has caused enormous exploitation of natural scares resources that are about to exhaust (Nejadi and Rahbar 2012) and has also created the issues like climate change and global warming by releasing a huge amount of carbon into the atmosphere (Sedjo and Sohngen 2012). Such distortion of the ecosystem and its services has negatively affected the economic capacity for sustainable development (Islam and Siwar 2010;Chandra et al. 2011;Nejadi and Rahbar 2012). Therefore, in recent decades, identifying the appropriate land use with good carrying capacity (Al-Mashreki et al. 2010;Aticho and Elias 2011;Nejadi and Rahbar 2012) along with suitable and nature-based solutions for climate change mitigation (Griscom et al. 2017;Koh et al. 2021) has been the prime concern of decision-makers. Forests are one of the most valuable natural resources that ght against climate change (Pache et al. 2021) by potentially reducing atmospheric carbon dioxide with the process of carbon sequestration (Vance 2018; Hou et al. 2019). Forests often act as carbon silos, due to their long life and a large volume of biomass, with net biological growth, offer huge volumes of carbon stored in their cells (Sedjo and Sohngen 2012). In spite of several intangible bene ts provided by the forests, the economic quanti cation of forest services has long been ignored because these values are di cult to measure and are not compatible with the conventional markets (Ninan and Inoue 2013).
The economic analysis of forest services such as carbon sequestration is crucial since it has the considerable economic potential to mitigate climate change (Raihan et al. 2019) and achieve emission reduction targets (Vass and Elofsson 2016). It also provides useful synergies for policymakers to accomplish the targeted objectives with the least net economic and social cost (Sampson and Sedjo 2020). There are a host of studies (De Jong et al. 2000;Ahtikoski et al. 2020;Hussainzad and Mohd Yusof 2020;Sampson and Sedjo 2020;Pache et al. 2021) that have estimated the economic value of forest carbon sequestration and payments for carbon sequestration in forests (Brooke 2009;Cho et al. 2018Cho et al. , 2019. Forest economic policies and incentives can largely in uence the process of carbon sequestration (Sedjo and Sohngen 2012) but there are various risks and uncertainties in the assessment of carbon sequestration related to carbon permanence, additionality, and leakage (Obersteiner et al. 2005;Sedjo and Sohngen 2012). There are also some issues associated with the payments for carbon sequestration i. e., choosing a viable reference with respect to which carbon storage is compared, variability across the places and the form taken by payments (Van Kooten and Johnston 2016; Gren and Zeleke 2016;Riviere and Caurla 2021). Regarding the economics of forest carbon sequestration, there is a huge scope of research that could lead to a general consensus on e cient valuation methods and policy instruments to promote it for climate change mitigation.
The current study sought to explore the scienti c literature on the "Economics of Forest Carbon Sequestration" (EFCS) by implementing a bibliographic network analysis as there are only a handful of studies that have carried out bibliometric mapping analysis on the economic aspects of forest carbon sequestration. This inference is derived when only 158 articles appeared after searching the keyword query-(("bibliometric analysis" OR "literature review" OR "citation analysis") AND ("vosviewer" OR "biblioshiny") AND ("forest" OR "biomass") AND ("carbon sequestration" OR "carbon capture" OR "carbon sink") AND ("economics" OR "economy" OR "economic analysis" OR "carbon assessment")) in the full-text. These 158 articles include bibliometric analysis, literature reviews, meta-analysis, scientometric mapping and cluster/content analysis. Publication fronts of some relevant literature reviews and meta-analyses are related to carbon storage estimation of forest ecosystem, the economics of forest carbon offsets, FCS cost studies, policy design for FCS, economics and policy of carbon sequestration, cost of carbon offsets, and economics of forest ecosystem carbon sinks (Richards and Stokes 2004;Van Kooten et al. 2004;Van Kooten and Sohngen 2007;Bangsund and Leistritz 2008;Van Kooten and Johnston 2016;Gren and Zeleke 2016;Sun and Liu 2020). Further, there are studies (Nardi et al. 2016;Uribe-Toril et al. 2019;Omoregbe et al. 2020;Yu et al. 2020;Wei et al. 2021) that have carried out bibliographic analysis also known as advanced statistic based science mapping by applying various bibliometric units let it be coauthorship, bibliographic coupling, co-citation, and co-occurrence analysis of key authors, documents, sources, references, institutions, countries, and keywords considering various research domains, i.e., payments for ecosystem services, carbon capture technologies for climate change mitigation, trends in carbon offset research, structure and evolution of forest research, etc. But to our knowledge, none of the studies has performed the bibliometric analysis on economics of forest carbon sequestration. The main aim of the present study is to demonstrate the structural evolution and dynamic demeanour of scienti c research in this eld by projecting the principle and relevant themes, documents, authors, journals, disciplines, countries, institutions, and keywords using Vosviewer and Biblioshiny software tools and Dimensions database over the period 2001 to 2021. This study will provide signi cant information to the scholars who want to have a scienti c overview of the multidimensional structure, trend pattern, fore and future fronts of researches under the current eld of study.

Bibliometric analysis methods
Bibliometric analysis, as one of the most rigorous practices, has been widely recognised for analysing the various aspects of published academic materials, including highly-cited documents, most in uential journals, countries, organizations, and to show a past and present structure of the concerned eld through citation, co-authorship, bibliographic coupling, keyword occurrences and cluster analysis (Zhang and Yuan 2019). This approach adopts three main bibliometric indicators to measure scienti c development: quantitative indicators (measure the productivity); qualitative indicators (measure the performance); and structural indicators (assess evolution and trend patterns of scienti c literature) (Biancolillo et al. 2020). There are several other reasons that attract researchers to adopt this method like research studies integrated with data is more relevant and scienti c than the subjective assessment (Nobanee et al. 2021), statistical and mathematical techniques are used to evaluate and predict the research status (Osareh 1996), and this technique guides in acquiring scienti c review overviews (van Eck and Waltman 2017).

Research software
There are numerous bibliometric software tools like CiteNetExplorer (van Eck and Waltman 2014), VOSviewer (van Eck and Waltman 2010), CiteSpace (Chen 2006), BibExcel (Persson, O., R. Danell 2009), and Biblioshiny (Aria and Cuccurullo 2017). Among them, VOSviewer (version 1.6.16, released on 25/Nov/2020) and Biblioshiny (Bibliometrix R tool, version 3.1.3, released on 25/May/2021) software have been incorporated in this study. VOSviewer software is a free Java application used for the creation, visualisation and evaluation of maps obtained from bibliometric data (Aria and Cuccurullo 2017). The new version of VOSviewer supports bibliometric data from Dimensions along with Web of Science, SCOPUS, and PubMed databases (www.vosviewer.com). On the other hand, the Biblioshiny app for bibliometrix redirects users to a web interface to perform visualization and bibliometric analysis without having any knowledge of coding (Aria and Cuccurullo 2017). The Biblioshiny app starts after installing the bibliometrix package (install.packages("bibliometrix")) in R Studio and running the biblioshiny () command in R Console (Waghmare 2021). Bibliometrix package applies to many bibliographic data from various databases like Web of Science, SCOPUS, Dimensions.ai, PubMed, and Cochrane (www.bibliometrix.org).

Data collection
This study has extracted bibliometric data from the Dimensions database, launched by the Digital Science Company in 2018 which have provided a new potential source of scholarly data as proved by many studies (Badke 2018;Orduña-Malea and Delgado López-Cózar 2018;Thelwall 2018;Martín-Martín et al. 2021). This innovative database re-imagines discovery and easy access to research materials by covering 119 million publications contextualized with more than 1.4 billion citations, policy documents, grants, clinical trials, linked patents and Altmetric attention openly accessible to users at app.dimensions.ai (www.dimensions.ai). An interesting feature of this bibliometric data source is the inclusion of new category lters based on 17 SDGs of the United Nations that has been incorporated in this study (Jackson 2020). All these features of Dimensions provide insights to explore new bibliographic ideas.
The retrieved research data were exported in .csv le format applying search criteria (keywords string: (("carbon sequestration" OR "carbon capture" OR "carbon sink" OR "carbon stock") AND ("forest" OR "forestry" OR "biomass") AND ("economics" OR "economic analysis" OR "economic valuation" OR "economic assessment" OR "carbon assessment" OR "carbon valuation" OR "opportunity cost" OR "carbon tax" OR "cost and bene t" OR "carbon pricing")) in Title and abstract (1,802 publications); Limit to articles

Main information and publication trend analysis
The basic bibliometric details on Economics of Forest Carbon Sequestration (EFCS), obtained from biblioshiny application, are presented in Table 1. There are 1,439 publications from 503 sources within the period 2001-2021. Average citations per article and average citations per year per document are 28.57 and 3.733 respectively. The total number of authors is 5,258 and references is 38,388. Authors per Documents index (3.65) is calculated as the total number of authors divided by the total documents. The Co-Authors per Document index (4.47) is the average number of co-authors per document. This index considers the author appearances whereas "authors per documents" counts an author only once, even if he is the author of more than one documents. That is why Authors per Document ≤ Co-Authors per Document. The Author Collaboration Index (3.97) is obtained as the ratio of the Authors of Multi-Authored Documents and Total Multi-Authored Documents (Elango and Rajendran 2012). In a sense, the Collaboration Index (CI) is a co-author per documents index obtained by only taking the multi-authored document set (www.bibliometrix.org).
The evolution of annual scienti c production is shown in Fig. 2  Panel (b) shows that SDG13: Climate Action has the largest number of documents (622 publications) that aims to "take urgent action to combat climate change and its impacts" (www.globalgoals.org). Such a huge number of publications under this category proves the importance of forests for combating climate change. 396 documents have been produced under SDG 7: Affordable and Clean Energy which main aim is to "ensure access to affordable, reliable, sustainable and modern energy for all" (www.globalgoals.org). SDG 15: Life on Land has 262 documents that aim to "protect, restore, and promote sustainable use of terrestrial ecosystems, sustainably manage forests, combat deserti cation, and halt biodiversity loss" (www.globalgoals.org). SDG 1: No Poverty; SDG 12: Responsible Consumption and Production; SDG 2: Zero Hunger have also signi cantly contributed in this eld followed by SDG 11, SDG 10, SDG 6, and SDG 8.

Documents bibliographic coupling analysis and top-cited articles
A network map of documents bibliographic coupling based on citations is generated through VOSviewer software in Fig. 4, showing the most in uential works of the highest citation. Of 1,439 publications, 908 meet the threshold while keeping the minimum number of citations to 5. Out of which 100 documents with the greatest total link strength are selected for the study, where the whole combined total link strength is 2785. According to Kessler (1963), bibliographic coupling occurs when two articles refer to common references. Two scholarly works are said to be more intellectually co-related when they have a large number of references in common indicated by a higher numerical value of link strength in this study. The size and colour of circles differ according to the referred common number of citations among the selected documents.
Cluster analysis determines cluster of objects that belongs to similar subject areas. In g. 4, ve clusters are found based on subject relatedness representing ve economic dimensions of studies under EFCS. The rst cluster is the largest one having 28 items (red nodes) followed by 23 items in the second cluster (green nodes), 20 items in the third cluster (blue nodes), 11 items in the fourth cluster (yellow nodes), and 6 items in the fth and smallest cluster (purple nodes). The assigned cluster names with their corresponding studies of highest citations along with their research fronts are shown in Table 2. Only those studies are included in each cluster that are relevant to the cluster themes. Cluster 1 is made of the studies related to the economics of forest carbon sequestration (FCS) mainly contributed by Corbera (2007)   Out of the most cited articles on the economics of forest carbon sequestration, the top 10 studies shorted by relevance as per the theme of the current study have been selected and shown in Table 3

Highly cited and most productive authors
The VOSviewer citation map of the most in uential authors extracted from the Dimensions bibliometric database is represented in Fig. 5. From the 5,250 total authors, the top 100 highly cited authors have been selected for the analysis, out of which 34 researchers met the threshold according to the relatedness of their number of co-authored publications. The graph is demonstrating 6 clusters and 116 total coauthorship links. The most cited author in each cluster representing the name of that cluster. For instance, cluster 1 is represented by Hertel, Dietrich having 567 citations. In the same way, cluster 2 is represented by Obersteiner, Michael securing 744 citations followed by Kriegler, Elmer with 402 citations, Popp, Alexander with 525 citations, Calvin, Katherine V with 1,305 citations, and Faaij, André P C with 644 citations. It is remarkable to observe from the map that no clusters have any overlap, evidencing relative independence of research groups under EFCS.

Citations analysis of Core Journals and Bradford's Law
From the analysis of subject categories and cluster analysis of articles, it is observed that studies on EFCS have multidisciplinary features and cover a wide range of research themes that have been published in a variety of journals with various research orientations. The source-citation analysis of core journals has been performed to understand the signi cance and collaboration of each speci c journal in creating the knowledge on EFCS. Journal density visualization map based on documents-weights is prepared through VOSviewer shown in Fig. 7. Of the 503 sources, producing 1,439 documents, 65 meet the thresholds keeping the minimum number of documents of a source to 5 and the minimum number of citations of a source to 5. The largest set of connected items are consist of 63 items. The colour schemes of the source density visualization map is of Rainbow type varying from red, yellow, green to blue colour where red corresponds to the highest document density and blue corresponds to the lowest document density. In other words, the colour of each node in the density visualization plat depends on the density of that node.
In this way, the density visualization pattern helps to understand the whole structure of the map by concentrating on the most relevant areas in the map. Hence, we can trace the most productive journals instinctively on the map. From Fig. 6, Forest policy and economics, Forest ecology and management, Forests, Journal of cleaner production and Journal of environmental management represent the core and most productive ve sources in the current eld of study. The top 10 leading journals with their frequencies, citations, H index, and start year are shown in Table 5   Bradford's law is a distribution pattern discovered by Samuel C. Bradford (1934) to measure the exponentially diminishing returns of reference searches in scholarly journals. Bradford claims that for a given subject eld "there are a few very productive periodicals, a larger number of more moderate producers, and a still larger number of constantly diminishing productivity" (Nash-Stewart et al. 2012). Under this law, if scienti c journals of a subject are arranged into three groups by the number of articles, each occupying about 1/3 rd of total articles, then the proportion of the number of journals in each group will be 1:n:n 2 (Table 6). These three groups are formally known as Zone 1 or core, Zone 2 or middle part, and Zone 3 or tail. Fig. 7, derived from Biblioshiny app, highlights the Zone 1 or "core journals" that are most frequently cited and the researchers in this eld are likely to have more interest in these journals. Zone 2 consists of the journals with average citations. Zone 3 re ects a long tail like picture of those journals that are considered as of marginal importance and very rarely cited (Potter 2010). Table 6 shows that Zone 1 includes 21 "core journals" which is 4.17% of the total 503 journals containing 486 articles. 3. 6 Co-authorship analysis of countries and institutions Regional collaboration, degree of communication, and research hotspots could be identi ed from the network visualization map of co-authorship occurrences among countries. The overlay visualization map of country co-authorship based on documents-weights in the eld of EFCS is represented in Fig. 8. Keeping the minimum number of documents and citations of a country to 5, of the 99 countries 51 meet the thresholds. For each of the 51 countries, the total strength of the co-authorship with other countries is analysed. The size of the frames represents the in uential countries according to the number of occurrences of documents. The distance between two nodes re ects the connection of their coauthorship link, and the thickness of the networking lines shows the strength of the link. The colour gradient is shown in the lower right corner of the map indicating the publication year of the author. The colour of the frame of a country is decided by the number of scores where blue represents the lowest score (the oldest or pioneer countries) and yellow represents the highest score (the latest or emerging countries).
The global distribution of documents published on EFCS is illustrated on the world map in Fig. 9 where the white spaces are representing no data. In the map, the colour gradient varies from red (highest publication) to green (lowest publication). Table 6 shows the top 10 countries of highest scienti c production with their citations and total link strength (TLS) values. From the overlay map (Fig. 8), world map ( Fig. 9) and Table 7, it is understood that the United States (US) is a pioneer country in this eld with the highest document production (336), citations (13315)   The overlay network map of co-authorship occurrences in institutes based on documents-weights are shown in Fig. 10.

Title's Keyword analysis and thematic evolution
The title's keyword analysis analyses the keywords that have been occurring in the title of the articles published in the concerned eld. Such kind of investigation is required to establish the research trend, ascertain the research gaps in the studies on EFCS, and discover the interesting elds and themes that need special attention of researchers. Word Cloud is an instrument for representing the word's frequency in terms of occurrences in a text body. This approach visualizes a bunch of words that appear in different font sizes depending on how frequently they have occurred in the source text. This lucrative and effective technique of keyword analysis brings forth an instant result of trendy research fronts.
To create the Word Cloud of keywords, R-Studio based bibliometrix tool-'Biblioshiny app' is used (Fig.  11). In the Biblioshiny app, under graphical parameters, the eld is selected as 'Titles', N-grams is selected as 'Bigrams', and the number of words is selected as '70'. The word occurrence measure is 'Frequency', shape-'Cardiod', font type-'Verdana', and text colour is 'Random Dark'. At last, the Ellipticity and Padding values are selected as 0.65 and 1 respectively. The Treemap of the top 20 most frequent keywords with their frequency and percentage of occurrences are illustrated in Fig. 12. The Word Cloud and Treemap both gures altogether depict that 'carbon sequestration' is the most used keyword with 201 frequency and 21% of occurrence followed by 'climate change', 'ecosystem services', 'carbon stocks', 'carbon capture', 'forest carbon', 'forest management', and 'carbon storage' that have usually occurred 9-4% of the times in the title of the articles, indicating the most highlighted, trendy, and signi cant part of the eld.
But at the same time, keywords like 'techno economic', 'economic analysis', 'economic assessment', 'socio-economic', and 'sequestration potential' have retained comparatively less percentage of occurrences that is about 3-2%. This analysis infers a huge research gap in the economic analysis of forest carbon sequestration where themes like carbon sequestration, climate change, and carbon capture are studied in the majority and the economic aspects of this eld are lacking attention. However, in recent times, the trend of themes has shu ed as delineated by Sankey diagram of the thematic evolution in Fig.13.
The thematic evolution technique identi es the changing paths of studies, evolutionary relationships as well as structures, contexts, and strength of emerging themes that appear over time. This method plays a very important role in portraying the degree and direction of eld development and also in forecasting the trends of the eld (Cobo et al. 2011). The Sankey diagram (Fig. 13) displaying thematic evolution is a kind of ow chart where the width of the arrow is proportional to the quantity ow. Each node in Fig. 13 corresponds to a topic and the width of the node is proportional to the frequency of keyword occurred under the theme. A thick node characterizes the relevance of that theme. Three-time zones (2001-2013; 2014-2017; 2018-2021)

Limitations Of The Study
This bibliometric literature on the EFCS replicate numerous facets of research achievements in the form of documents, journals, countries, institutions, and keywords analysis, but also suffers from some thesis, 4. References analysis is not included because it is not possible to perform all the bibliometric measures in a single study, 5. Many articles fully devoted to the concept of EFCS are missing from the study because the same terminology does not explicitly exist in the titles and abstracts of those articles as used in the keyword search string of the current study, and 6. This study does not perform deeper investigation, for instance, methods/models and the signi cant outcomes, which need to be further integrated.

Discussion, Conclusion And Future Research Recommendation
In this fast-growing complex world, our environment and ecosystem are facing lots of side-effects of rapid and unsustainable development like GHG emissions, heat waves, and climate change. To protect our earth from the climate crisis, targets have been set to limit global warming to well below 2 0 C, preferably to 1.5 0 C, compared to pre-industrial levels. To achieve this target, global and national policymakers are inclined to adopt a low carbon cost-effective technology and forest carbon sequestration (FCS) is one of them. Analysing the economic aspects of FCS is essential for providing incentives to the national governments, landowners, and stakeholders which further helps to achieve green economic growth and sustainable development. Nevertheless, the bibliometric studies speci cally related to the economic aspects of FCS are almost negligible, the ndings of the current study are consistent with several other existing bibliometric literature related to forest carbon sequestration, evolution of forest research, carbon capture, and carbon offsets. Many studies (Nardi et al. 2016;Aznar-Sánchez et al. 2018;Uribe-Toril et al. 2019;Nobanee et al. 2021;Wang et al. 2021) have found the similar most frequent keywords like 'Climate Change', 'Ecosystem Services', 'Carbon Sequestration', 'Forest Management, 'Economics', and 'Economic Valuation' as the present study. The main subject categories of this multidisciplinary eld are-Environmental Sciences, Agriculture, Biological Sciences, and Economics that is also veri ed by the following studies (Aznar-Sánchez et al., 2018;Huang et al., 2020;Uribe-Toril et al., 2019;Wang et al., 2021;Wei et al., 2021).

Figure 8
Overlay visualization map of country co-authorship Figure 9 Global distribution of documents among countries Thematic evolution based on word occurrences