Statistics and Analysis of Global Change Research Data Publishing & Sharing (2014-2017)

Shi, R. X.^*  Liu, C.  Ma, J. H.  Wang, Z. X.

Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China

Abstract: As the first case in global change research data publishing and repository in World Data System (WDS) of International Council of Sciences (ICSU) and China GEO, 343 datasets in 17 issues were published in GCdataPR (Global Change Research Data Publishing & Repository) during 2014-2017. In which, 11 datasets were global scale, 23 datasets are trans-continental, 248 dataset (72.30%of total) cover Asia area. There were 567 authors from 11 countries; most of them (94.89%) came from China, mainly from the Chinese Academy of Sciences, and universities from the Ministry of Education of P. R. China. The online data size is 1,028.13 GB (209.10 GB compressed). More than 880,800 visitors and 36,800 computer IP users from 76 countries over the world downloaded 125,900 data files openly and freely with the data size of 2,675.70 GB during 2014-2017. China was the largest user group (32,749, 89%), followed by the United States, Australia, Japan, and Malaysia. The GCdataPR was on the top 50 of Big Data Products, Services and Solutions of China (the only one from scientific research and education) in 2016 and searchable list of Data Citation Index of Clarivate Analytics since 2016.

Keywords: global change; research data; data publishing; 2014-2017

1 Introduction

Since the National Global Change Research Program^[1–3] launched, the global change research datasets have been continuously produced. In order to promote the data publishing and openly available to be re-used, the Institute of Geographic Sciences and Natural Resources Research (IGSNRR), Chinese Academy of Sciences (CAS) and the Geographical Society of China (GSC) officially launched the Global Change Scientific Research Data Publishing & Repository (GCdataPR) in June 2014. There were 343 datasets in 17 issues published from 2014 to 2017.

According to the Guidelines of Global Change Scientific Research Data Publishing & Repository^[4], the annual report is conducted in two ways. The first is data publishing and sharing rankings, which is released in mid-year and the second is a comprehensive summary, which is released by end of each year. The GCdataPR 2017 Rankings was released in June 2017^[5–6] and this article is the summary of the annual datasets repository report 2017, which covers from June 1, 2014 to December 20, 2017.

2 Statistics and Analysis on Dataset Published

2.1 Datasets Published

In total, 343 datasets were published in 17 issues from 2014-2017. Of these, 36 datasets published in 2 issues in 2014 and 34 datasets in 2 issues in 2015 separately, 190 datasets in 9 issues were published in 2016. While, in 2017, in conjunction with the Journal of Global Change Data and Discovery, 83 datasets in 4 issues were published (Table 1). There are 394,026 data files in total among the 343 datasets, which were compressed into 1,226 data file packages. The compressed data file packaging rate is 321.39. The total data size is 1,028.13 GB, 209.10 GB compressed; the data compressing rate is 4.92.

Table 1  Statistics of datasets published and archived (2014‒2017)

2.2 Geographical Regions Covered by the Datasets

There are 11 datasets are in the global scale among the 343 datasets, which is 3.21% of the total; 23 datasets covering the cross-continental region, which is 6.71 % of the total. Most datasets cover Asia area (248 datasets), which is 72.30% of the total. Of these, 148 datasets cover China, accounting for 59.68% of the Asian and 43.15% of the total. There are 21 data sets covering Europe, 6.12 % of the total. There are few datasets cover Oceania, Africa and Latin America, 3 for Oceania and Africa and 4 for Latin America, respectively. 16 datasets cover North America, accounting for 4.66 % of total. And there are 11 datasets covering the Polar Regions, accounting for 3.21% of the total. In addition, there are 3 datasets dealing with data the new technologies (software and video, Table 2)

2.3 Datasets Published by Disciplines

The datasets published covered a wide range of spectrum in the global change sciences (Table 3). There were 228 datasets about land (66.47 %); 108 datasets about ocean (including polar regions, coastal area, and islands; 31.49%); 4 datasets about earth-science-related arts (1.17 %); 3 datasets about technology (0.87 %). Among the land datasets (228), there were 40 datasets about water in land (river, lake, and wetland) and 24 datasets about land cover and land use. The largest section of land datasets was from terrestrial ecosystems, totaling 79 datasets.

2.4  Data Levels

All datasets were identified into 0-5 levels, according to their stages in the development procedures. Specifically：

Level 0 (L0)：Raw data or signal from sensor or observer;

Level 1 (L1)：Output of L0 after geometric and radiometric correction;

Level 2 (L2)：Output of L1 integrated with new intelligence;

Level 3 (L3)：Output of L1/L2 integrated with new intelligence;

Level 4 (L4)：Output of L1/L2/L3 integrated with new intelligence;

Level 5 (L5)：Output of L1/L2/L3/L4 data integrated with new intelligence; usually a time-serial, or global scale.

Based on above criteria, 343 datasets were categorized into 3 production levels (Table 4): 72.89% of the datasets were in level 2; 24.49% of them were in level 3, and only 2.62% of the datasets were in level 4.

3 Dataset Author(s)

3.1 Dataset Author

The 343 datasets were developed by 567 authors from 11 countries, including China, Japan, Kenya, Thailand, etc.

Most datasets were contributed by Chinese scholars (94.89%). Only Six datasets (1.75%) were published by non-Chinese authors alone，16 datasets (4.66%) were results of the international cooperation (Table 5).

3.2 Dataset Author Groups

Among the 343 dataset and 567 authors，only 7.29% of the datasets were developed by a single author, 74.93% of the datasets were developed by a group consisted of 2-5 persons, and 16.33% by a group consisted of 6-10 persons. There were 5 datasets from the groups of over 10 persons, and the largest group was composed of 33 members (Table 6).

3.3 Statistics of Chinese Authors by Affiliation and Region

3.3.1 Datasets Authors by Affiliation (Institutes or Universities)

Table 7 presented the affiliations of datasets authors. Most of the authors are from Chinese Academy of Sciences and universities belong to the Ministry of Education of P. R. China (Table 7).

3.3.2 Chinese Authors by Regions（Province, Municipality, Autonomous Region）

The current distribution of dataset authors was quite uneven: authors from Beijing published 256 datasets, obviously the lion’s share of the total publication; the following 5 provinces, Shanghai, Sichuan, Henan, Gansu, Jiangsu contributed 10-20 datasets, respectively; there were no datasets from six regions, such as Ningxia, Qinghai, Guizhou, Hong Kong, Macao, and Taiwan (Table 8).

Table 8   Statistics of Chinese authors by region

3.4 Statistics of Datasets by Founding Agencies

With or without foundations (Table 9): Of all the published datasets, 84.55% of them were founded, 15.45% of them were conducted by self support. Among the with-founding group, 47.23% of the datasets were funded by a single funding project; 37.32% of the datasets were funded by two or more funding projects, those datasets were usually large, broad-covered, time-serial datasets.

Foundation details (Table 10): Of all 343 datasets from 532 projects (programs) , 139 projects (26.13%) were from Ministry of Science and Technology of P. R. China; 123 projects (23.12%) were from Chinese Academy of Sciences and 118 projects (22.18%) were from Natural Science Foundation of China.

Table 10  Statistics of funds supporting datasets development

4  Association of Datasets, Research Papers, and Data Papers

In order to promote the informed research data re-use, products in three dimensions have been considered: research datasets, research data paper and original research paper^[7]. GCdataPR was the first case in World Data System (WDS) of International Council of Sciences (ICSU) and China GEO^[8].

The publication of research data through the platform of Global Change Research Data Publishing and Repository (GCdataPR, http://www.geodoi.ac.cn) gives authors a new choice to protect their data and enjoy relevant credit. By the end of 2017, 37 academic journals have participated GCdataPR partner team, they can enhance their visibility through recommendation of research data from the papers published in their journals. GCdataPR was on the top 50 of Big Data Products, Services and Solutions of China (the only one from scientific research and education) in 2016^[9] and searchable list of Data Citation Index of Clarivate Analytics since 2016.

Development of research data can be very complex and may not be fully instructed in original research papers. Therefore, it is necessary to provide more information to new users in order to apply the data in new research fields. This is so called data paper, published through Journal of Global Change Data and Discovery.

To give potential data users more insight about research data, the online hot link for each dataset to original research paper is provided at the GCdataPR platform. This means that data can be re-used on a transparent background, and credits and responsibilities of each part can be clarified in the very beginning.

It is worth noting that the above three steps have not been taken shoulder by shoulder. GCdataPR was launched in 2014, part of its data papers were published in Journal of Geography. This lag of data paper publication remained until March, 2017, when the new journal delivered its first issue. With the completion of trinity, more data papers will be published through the new journal. Table 11 gives a brief for this trend.

5 Statistics of Data Sharing

5.1 Datasets

Table 12 presented the major static items of data sharing during 2014-2017. More than 880,000 users visited the system, accumulatively. Through over 30,000 IPs, users have downloaded more than 2TB datasets in 120,000 times. The data users (IP) and the data file downloaded increased year by year (Figure 1, Figure 2).

Table 12  Statistics of data sharing through the GCdataPR

Figure 1  Statistics of annual added and cumulative data users of GCdataPR (2014–2017)

Figure 2  Yearly and cumulative data file downloaded from GCdataPR (2014-2017)

5.2 Data Users by Countries

Data users from 76 countries all over the world have downloaded the datasets. Most of the users were from China (32,749, 89%), followed by the United States, Australia, Japan, and Malaysia (Figure 3).

Figure 3  Map of the GcdataPR data users (IP)

6 Discussion

Global change research data is comprehensive and inter-disciplinary, just like Global change research. There is no doubt that the datasets and the data paper publishing together with the citation of the original research paper, could provide the scientific community with a new possibility to facilitate data quality assurance and informed research data re-use.

However, at its preliminary and exploratory stage, several key issues should be tackled further in a wide range of fields, such as data publishing, archiving, disseminating, sharing, application, citation, measurement of impact, etc. The most marked issues are: scientific community as a whole has not given enough credit to research data published; lack of enforcement of intellectual property protection for research data; inconsistency of data citation, the data content may need cover more regions and the data authors may expand to more communities.

Although the IGSNRR/CAS, SGC try hard to enhance the capacity building and communication in research communities and universities under the support of Chinese Academy of Sciences, the awareness of research data publishing should be more involved, especially from founding agencies; research institutes, universities. The due credit should be given to the each of stakeholders related to the published datasets.

There are still many confusing issues about data intellectual property protection among the scientific communities and the society. For instance, what makes up an innovative dataset and what should be protected by law? How to balance the protection and sharing? It is valuable to have such course as a basic education in universities about the data intellectual property.

Unlike paper citation, there are no common understanding the guidelines and standards for data citation. Usually, the data used in a paper may be specifically cited in terms of “acknowledge”, or in “data and method”, or just overlooked. In short, there was no such a tradition as paper citation before. That was history. In our new system, the data users are requested to formally cite both dataset according to the dataset citation format and the data paper according to the data paper citation format.

References

[1]    Future Earth (2013). Future earth initial design: report of the transition team [R]. Paris: International Council for Science (ICSU).

[2]    Experts Group of National Major Research Program on Global Change Research. Global change strategy research report of China [R]. 2009.

[3]    USGCRP. The national global change research plan 2012–2021: a strategic plan for the U.S. global change research program [R]. 2012.

[4]    Editorial office of Journal of Global Change Data & Discovery. Guidelines of global change research data publishing and repository [J]. Journal of Global Change Data & Discovery, 2017, 1(3): 253-261. DOI: 10.3974/geodb.2017.03.01.

[5]    Geographical Society of China. Global change research data publishing & sharing rankings (Top 10) [J]. Journal of Global Change Data & Discovery, 2017, 1(2): 249-251. DOI: 10.3974/geodp.2017.02.23.

[6]    Jiang, D., Song, X. F., Zhang, G. Y. A new milestone of scientific data sharing in China [R]. Journal of Global Change Data and Discovery, 2017, 1(2): 246-248. DOI: 10.3974/geodp.2017.02.22.

[7]    Liu, C. Global change research data publishing and repository [J]. Acta Geographica Sinica, 2014, 69(Sup): 3-11.

[8]    Liu, C., Guo, H. D., Uhlir, P. F., et al. GCdataPR: infrastructure for data publishing repository & sharing in/for/with developing countries [J]. Journal of Global Change Data & Discovery, 2017, 1(1): 3-11. DOI: 10.3974/geodp.2017.01.02.

[9]    Shi, R. X., Zhu, Y. Q., Jiang, D., et al. GCdataPR: an outstanding case in “Excellent Products, Services and Applications of Big Data in China” [R]. Journal of Global Change Data & Discovery, 2017, 1(2): 245. DOI: 10.3974/geodp.2017.02.21.