2016 User Survey - Europe

General Remarks

This part of the survey only considers publications registered at one (or more) European Photon and Neutron facilities, where publication dates were in 2010-2016. All data have been obtained from the institutional databases in December 2016. All other details can be found in the global survey.

The facilities

We collected publication data from 15 different facilities for the years 2010-2016 (where available). Please resist the temptation to compare facilities by number of publications or users!

The idea behind these surveys is to document the commonalities across facilities and scientific disciplines, not to establish a metric. All facilities are in a very different state, and experimental setup requirements can be extremely different. For a example a standard MX experiment can be done within minutes (including structure determination), whereas a complex setup at a Neutron source or FEL can easily take more than just a few days.

Some global figures

Number of facilities included: 15
Number of Neutron facilities: 5
Number of Photon facilities: 10
Number of (unique) publications (DOIs): 36030
Number of (unique) users (authors): 84157
Countries of users home institutions: 117

Total number of unique users for individual years:

year 2010 2011 2012 2013 2014 2015 2016 Sum 2010-2016
users 17933 19125 20855 22803 24358 24937 23756 153767 84157
publications 4661 4863 5005 5514 5650 5549 4788 36030 36030

Number of users are continuously increasing. The counts for 2016 are somewhat lower simply because recording of publications takes some time and are presumably not very complete yet for 2016.

In previous years, the survey was covering a timespan of two years to smooth out extended shutdown periods. The lengthy publication processes tend to smooth out such irregularities, so that stats on smaller timeframes are much less affected. To make comparison with previous surveys easier, the corresponding 2 year figures for publications and authors:

year 2010-11 2011-12 2012-13 2013-14 2014-15 2015-16 2010-2016
users 30339 32657 35756 38675 40592 40240 84157
publications 9524 9868 10519 11164 11199 10337 36030

In the survey for 2013-2014 we found a total of 41665 and for the survey in 2012 a total of 33025 unique users. Despite the small variations in facilities included in the surveys, the number are remarkably similar to the numbers obtained from publications, which makes us confident that the figures are indeed fairly accurate and sensible.

brief overview of publications and users per facility

# Facility Type P|N Country Region DOIs Public # Publications # Users U/P Map
1 ALBA Synchrotron P Spain Europe yes yes 368 1802 4.9 g
2 ANKA Synchrotron P Germany Europe yes yes 348 1553 4.5 g
3 DESY Synchrotron/FEL P Germany Europe yes yes 2305 8508 3.7 g
4 DLS Synchrotron P UK Europe yes yes 4683 17843 3.8 g
5 ELETTRA Synchrotron/FEL P Italy Europe yes yes 2436 7621 3.1 g
6 ESRF Synchrotron P France Europe yes yes 13342 40213 3.0 g
7 FRM-II Fission Reactor N Germany Europe yes yes 1191 3769 3.2 g
8 HZB Ber II Fission Reactor N Germany Europe yes yes 574 1817 3.2 g
9 HZB Bessy II Synchrotron P Germany Europe yes yes 2640 9272 3.5 g
10 ILL Fission reactor N France Europe yes yes 3470 10610 3.1 g
11 ISIS pulsed neutron and muon source N UK Europe yes yes 2871 8442 2.9 g
12 MAX Synchrotron P Sweden Europe no yes 868 3220 3.7 g
13 PSI SINQ Spallation Neutron Source N Switzerland Europe yes no 861 2590 3.0 g
14 PSI SLS Synchrotron P Switzerland Europe yes no 3597 14510 4.0 g
15 SOLEIL Synchrotron P France Europe no yes 1340 6398 4.8 g

User counts

The procedure is rather simple. For each of the European PaN facilities we collected all relevant publications (DOIs) published in 2010-2016 from their respective publication database. For each DOI we retrieved the available authors, affiliations and identifiers, utilizing e.g. CrossRef and Orcid APIs. Unfortunately the coverage of researchers with uniq ID is not very high, reserve retrieval of Orcid IDs for a DOI has a low success rate and at least some authors obviously have multiple IDs.

In the end, we considered authors with identical family names and identical given names (all have to match) the same author. It's possible that we match authors which are not the same person. It's also possible, that we miss many matches simply because authors are not very consistent with 2nd to nth given names (or initials). However, the resulting figures look so similar compared to previous surveys, that the assumption that missing matches and mismatches do not affect the statistics significantly appears well justified.

Of course we know that almost all facilities have common users/authors simply because almost all two facilities have DOIs jointly registered at both places. So the main conclusion, that users are tightly connected across all facilities. The circos diagram gives an idea about common DOIs. To put it on scale, ESRF and DLS share a total of 726 DOIs.

Circos diagram of DOIs commonly registered at facilities.

Users in Common

Users Geography

The publication related information made it possible to obtain the home institution and geographical location for 94% of the authors/users. This is a much higher fraction of users than for previous surveys. The gross picture is however virtually identical to those from previous surveys, which again confirms that the publication based data and the user office based data provide a nearly identical view on the users of the Photon and Neutron facilities in Europe.

Migration Background

Earlier surveys have shown that the user community of the Photon and Neutron facilities is strongly changing within a two years timeframe. Based on the publication data, we can get a more detailed picture. Looking at the users/authors of publications in individual years, we can get an estimate how many of the users are actually returning users:

Year 2010 2011 2012 2013 2014 2015 2016
2010 100% 37.4% 36.5% 35.3% 32.4% 31.7% 28.1%
2011 35.1% 100% 38.2% 36.9% 35.1% 32.6% 29.1%
2012 31.4% 35.1% 100% 37.8% 36.1% 33.6% 29.2%
2013 27.7% 30.9% 34.6% 100% 37.2% 34.8% 30.6%
2014 23.8% 27.6% 30.9% 34.8% 100% 35.7% 31.6%
2015 22.8% 25.0% 28.1% 31.8% 34.8% 100% 33.8%
2016 21.2% 23.4% 25.6% 29.4% 32.4% 35.5% 100%

So only 35.1% of the users in 2011 also showed up in 2010 (and 37.4% of the users in 2010 also showed up in 2011). In general, more than 60% of the users are kind of new users every year, which is very consistent with earlier results. Looking at individual facilities, one gets an idea how many users showed up for the first time at this facility or at any facility at all:

Facility Unknown at facility in 2015 Unknown at facility in 2010-15 Unknown at any facility in 2015 Unknown at any facility in 2010-15
alba 77.7% 66.6% 56.4% 34.0%
anka 82.1% 68.9% 63.8% 43.7%
desy 72.7% 51.7% 53.4% 29.2%
dls 73.1% 55.5% 60.7% 38.6%
elettra 67.5% 45.8% 51.5% 29.2%
esrf 72.2% 46.8% 59.7% 35.7%
frm2 64.3% 43.9% 45.2% 22.4%
hzb-ber 65.8% 43.9% 40.7% 20.8%
hzb-bessy 66.0% 46.7% 50.0% 31.0%
ill 71.5% 46.5% 52.2% 28.3%
isis 75.1% 54.6% 56.4% 33.7%
psi-sinq 67.8% 47.6% 42.5% 20.2%
psi-sls 72.3% 53.5% 57.0% 35.6%
soleil 83.4% 64.5% 61.8% 35.5%

ESRF is a typical example: 46.8% of the users haven't show up any time earlier and 35.7% of the users are not being found in the databases of even a single facility. So half of the users typically show up for the first time at a facility and one third appear to be real newbies to experiments at large scale user facilities, which gives an indication about the support level expected/required from user facilities.

Collaboration Profile

Most of the European PaN facilities are national facilities, but offering scientific opportunities on a global scale based on scientific excellence rather than location of researchers. The significant fraction of users shared between facilities is a good incentive for the facilities to work on interoperable data infrastructures facilitating data intensive research across the PaN facilities. Most experiments are of course pursued in large and international collaborations, which makes data sharing across and beyond the collaborations a pre-requisite for efficient data analysis.

This survey allows to gain some more insights into the broadness of the collaborations. The DOIs allow to extract all authors located in a specific country. Based on these DOIs we can collect all co-authors of publications with at least one authors located in Italy (UK, Spain,...).

This actually gives a good idea about the collaboration profile of researcher located in Italy (UK, Spain...). Researchers are of course always embedded in collaborations with members of the same lab, so it's little suprising that e.g. researcher working in an Italian lab are predominantly collaborating with researchers located in the same or closely related lab. However, in all cases, even for countries with several user facilities like France or Germany, more than 50% of the collaborators (co-authors) are coming from abroad.

The table (heatmap) gives an overview of the collaborating authors across different countries (for a selection of countries). For example, 18.8% of the Austrian PaN users co-author publication with German PaN researchers (but only 1.5% of the German Pan-users with Austrian researchers, simply because there are many more German PaN-users highlighting the number of facilities in Germany). The tight network of collaborations across countries and continents is remarkable.

Having a closer look at collaboration for individual countries shows always a very similar picture: typically 30-40% of the collaborations are within a country. The remaining 60-70% are widely distributed across different countries. Little surprising, the countries running one or more user facilities have a much more significant footprint. The only country with more in-country than foreign collaborations are the USA.


Collaboration profile for UK (only taking account European PaN facilities)

Collaboration profile for Portugal (only taking account European PaN facilities)

Collaboration profile for USA (all facilities)

Collaboration profile for Japan (all facilities)

Publishing

Since this survey is based on the DOIs registered at the PaN facilities, it's very simple to obtain information about publishers and journals for example from CrossRef or Thomson Reuters Web of Science.

Publishers

The distribution of publications across publishers doesn't bear any surprises. The distribution is dominated by a handful of publishers, and the top 20 publishers (out of more than 200) publish way more than 90% of all PaN publications. The numbers don't have to very accurate, and not all publishing groups are represented as a single publisher (e.g. Springer and Nature), however the figure gives a good idea and shows that the regional variations tend to be small - with two exceptions. Seemingly the Asian-Pacific researchers tend to publish much more in journals of the Royal Society of Chemistry, whereas the American Researcher more strongly favour the American Chemical Society (less of a surprise).


Distribution of PaN science publication across publishers for all facilities, the European, American and Asian-Pacific facilities. All numbers are in %.
 

Journals

Looking at individual journals, the distribution of publications from PaN researchers is much more wide-spread. The most popular journal with roughly 5% of all publications (from all facilities in this survey) is Physical Review B. Following is a long list of journals from various fields with 1-2% of the publications, the figures below show the 30 most frequently named journals. The wide variety of journals reflects the diversity of scientific questions investigated at the PaN facilities, and the large number of scientific communities utilizing our instruments.


Distribution of PaN science publication across journals for the European, American and Asian-Pacific facilities. The numbers are in percent relative to the total number of publications from any of the PaN facilities.

Relative popularity of journals for researchers of the European, American and Asian-Pacific facilities. In shows that for example the journal Physical Review B is almost identically popular in all three regions; the journal Biochemistry is much more popular in the American region than in Europe or Asia-Pacific regions.

Impact

Impact is a terrible metric for scientific success, in particular when a wide spectrum of different scientific fields is considered; high impact in structural biology is not really comparable to high impact in medicine or condensed matter. But it's a metric which is available for most journals used to publish results from PaN scientific results. The following two graphs simply sort all publications from European, American, and Asian-Pacific PaN facilities according to the impact factor of the journal used for the publications. Impact factor where collected from various sources and might not always be highly accurate and are current impact factors rather than impact factor at the date of publication.


Absolute number of publications in journals with a certain impact factor. Numbers are slightly higher than total number of publication due to publications registered in various facilities.

Absolute number of publications versus the number of citations a publication has received. 280 publications, which had been cited more than 100 times have been omitted from the plot (h-index is 133). Only publications with a publication date in 2010 have been used. The plot indicates that the impact of the PaN publications is considerably higher than the corresponding journal impact factor.

Research Subjects

Research, publications or journals are usually categorized or indexed in some way. Also authors usually provide some keywords or tags which they believe to most appropriately classify the work. Authors keys are not normalized in any way, and obviously are a particularly poor choice to classify research at the PaN facilities. Scopus, Web of Science and publishers classify journals and publications using a well defined vocabulary. Unfortunately it seems that no two of them actually use a similar dictionary. Looking at the 10 most popular categories obtained from various sources (for PaN registered publication), the diversity in terms and ranks is surprising:

Rank WoS Category WoS Research Area CrossRef Subjects Altmetric Subjects Scopus Subject Publisher Subjects
1 materials science, multidisciplinary physics condensed matter physics chemistry physical sciences chemical sciences
2 chemistry, physical chemistry electronic, optical and magnetic materials science life sciences biological sciences
3 biochemistry & molecular biology materials science chemistry(all) biochemistry biochemistry, genetics and molecular biology physical sciences
4 physics, condensed matter biochemistry & molecular biology physics and astronomy(all) molecularbiology chemistry multidisciplinary
5 chemistry, multidisciplinary science & technology - other topics physical and theoretical chemistry biology physics and astronomy engineering
6 physics, applied biophysics materials science(all) biophysics materials science medical and health sciences
7 multidisciplinary sciences crystallography biochemistry medicine chemical engineering inorganic chemistry
8 biophysics instruments & instrumentation molecular biology chemistrytechniquesanalytical general biochemistry and cell biology
9 crystallography cell biology materials chemistry biotechnology health sciences other chemical sciences
10 nanoscience & nanotechnology optics surfaces, coatings and films radiology medicine medicinal and biomolecular chemistry

Despite the diversity in classifications, it's rather obvious that Material Science is the most frequently named research field for research at the PaN facilities, followed by Life Science and more fundamental research in Physic and Chemistry.

 

Funding

Records from Thomson Reuter and recent records from CrossRef contain some funding information. The Thomson Reuter data seem to be mostly the plain information from the publication as provided by the authors. Unfortunately authors do not obey to any particular form or standards acknowledging funding which results in a wide variety of different names for the same funding agency. There are for example literally hundreds of different entries for the "Deutsche Forschungsgemeinschaft". Apart from misspellings. differences in hyphenation, and translations (e.g. German Research Foundation, Science Foundation, Funding Agency, Reseach Council, Science Council all refer to same funding agency) there are all kind of innovative abbreviations and combinations with specific sub-programs and (meaningless) additions.

The CrossRef seem to be considerably more homogeneous. For example the Thomson Reuter data for 2016 contain more than 100 different variants for ESPRC whereas CrossRef lists only 10 different versions of the same thing; in particular poor and semi-translations like "German Minister für Bildung und Forschung" are still escaping. There are also quite a number of entries like "DESY and HZB" which seemingly could not be matched with entries in CrossRefs Funder Registry. Several entries list Synchrotrons as funders. In most cases it's actually an acknowledgement of beamtime provisioning (parts of this research were carried out at the light source [labs]), provisioning of other resources ( We acknowledge the computing resources that are provided by [labs], technical support (acknowledged for their success in providing the data-analysis environment or just ideal support (We thank the [labs] [staff|directorate|...] for their strong support. All this support and resource provisioning is highly valuable (and the monetary equivalent easily exceeds the contributions from funding agencies), but it's not exactly funding data. It might help to track contributions of PaN user facilities to the research, which however doesn't seem to work too well considering the small number of entries for the PaN facilities (and only DESY, ESRF, PSI of the European PaN facilities seem to be listed in the CrossRef Funder registry).

The pie chart below lists some of the major funders as found in the publications registered at the European PaN facilities and only for 2016 assuming that the most recent data have the highest quality. It just sums the occurrences of funder in the records, and of course it's completely impossible to quantify any contributions. Despite the extreme limitations of the data, it's interesting to see that for example US NSF and DOE or Chinese funding agencies have quite a contribution to European research, highlighting again the strong networking of PaN research.

PaN facility # of occurrences
EU 620
DOE 90
ESRF 67
DLS 24
PSI 18
ELETTRA 13
DESY 11
ILL 10
HZB|BESSY 7
ISIS 4
SOLEIL 4
ALBA 2

EU and DOE shown for comparison.