changes open science in high-energy physics - salvatore mele at opencon

Change Alley: Open Science in High-Energy Physicsa.k.a. what’s impossible and what’s not

����

November 14th, 2015�OpenCon 2015 - BrusselsSalvatore.Mele@CERN.ch

HEP – High Energy Physics (crawling Wikipedia)

What is the world made of ?

International Catalog of World Coins

How does stuff work ?

flickr.com/photos/paulm/2253776428/

International co-operation

LHC 10’000+ scientists+engineers, 113 countries, 20+ years

27 Km, -271.25°C, 99.999999% of speed of light

Four “detectors”

Large as a cathedral

flickr.com/photos/the_yes_man/4012514457

100 million “sensors”, 40 million pictures/second

More than 100PB(=100’000TB) on tape at CERN

Welcome to CERN

Discovery of the Higgs boson

article

2899 authors

references

90% of HEP articles: 1-5 authors (mostly theorists)

Change Alley: Open Science in High-Energy Physicsa.k.a. what’s impossible and what’s not

����

November 14th, 2015�OpenCon 2015 - BrusselsSalvatore.Mele@CERN.ch

L3 detector at CERN LEP accelerator

Head of Open Access at CERN

Wikipedia

Change Alley: Open Science in High-Energy Physicsa.k.a. what’s impossible and what’s not

����

November 14th, 2015�OpenCon 2015 - BrusselsSalvatore.Mele@CERN.ch

flickr.com/photos/londonmatt/3163571645

flickr.com/photos/funky64/3925955346

impossible �

Observer, Alamy

Wikipedia

Scholarly communication� would have been impossible �

without scientific journals

Observer, Alamy

smaismrmilmepoetaleumibunenugttauiras

Altissimum planetam tergeminum observavi

Once upon a time, when air-mail was fast…

…HEP scientists wrote papers…

…passed them through the cyclostyle/mimeograph…

Carnegie Library of Pittsburgh

…then mailed them to journals AND colleagues…

…other scientists read these PREPRINTS…

…libraries catalogued these PREPRINTS…

… into Open Access repositories.

T. Berners-Lee, CERN, ’91: the web is born

P. Ginsparg ’91: “Preprints on the internet?” arXiv.org

97% of HEP journals’ content is in arXiv.org

Effects of global (green) Open Access in HEP �

Open Access accelerates Science!

Scientific dialogue on subject repositories

97% of HEP journals content is OA as preprints!

Journals no longer have a communication role!Open Access subject repositories accelerate Science!

Embargos would seriously hamper Science!

HEP articles also available OA!

Years !

Cita

tatio

ns!

0 ! 1 ! 2 ! 3 ! 8 !-1 ! 4! 5 ! 6 ! 7 !

Only published !

Gentil-Beccot, Mele, Brooks arXiv:0906.5418

Is impossible to convert �existing journals to Open Access �limiting the use of fresh money, �

and with no burden for researchers

flickr.com/photos/ipalatin/6031936991

Observer, Alamy

���

� ���

SCOAP3.orgSponsoring Consortium for Open Access

Publishing in Particle Physics

Re-use the CERN model…

…of international co-operation…

build a global

partnership

Re-use money …

…spent by libraries for subscriptions…

…and liaise with Funding Agencies…

…to pay peer-review/publishing services…

…and not for content...

…which is “mostly open” anyhow!

Keep scientists happy !

Sweden 0,8%Mexico 0,8%

Taiwan 0,8%Portugal 0,9%

Netherlands 0,9%

Iran 0,9%Israel 1,0%

Poland 1,3%Switzerland 1,4%

Korea 1,8%CERN 2,0%

India 2,6%Brazil 2,6%

Canada 2,7%

Spain 2,9%

Russia 3,4%

France 3,8%

China 5,3%

United Kingdom 6,7%

Italy 6,9%

Japan 7,2%

Germany 9,1%

United States 24,9%

Other Countries 9,3%

Cern Scientific Information Service

Distribution of HEP publications, average 2005-2006

J. Krause et al. CERN-OPEN-2007-014

Estimated cost: 5M€/year fairly distributed:each country contributes share of HEP publications

Took some time to get organized

Publisher Journal

Nuclear Physics B

Physics Letters B

Advances in High Energy Physics

Chinese Physics C

Journal of Cosmology and Astroparticle Physics

New Journal of Physics

Acta Physica Polonica B

Progress of Theoretical and Experimental Physics

European Physical Journal C

Journal of High Energy Physics

Start SCOAP3

Publisher Journal articles

Nuclear Physics B 605

Physics Letters B 1’659

Advances in High Energy Physics 312

Chinese Physics C 44

Journal of Cosmology and Astroparticle Physics 414

New Journal of Physics 17

Acta Physica Polonica B 33

Progress of Theoretical and Experimental Physics 148

European Physical Journal C 1’045

Journal of High Energy Physics 3’839

Articles as of November 13th 2015:Share of all HEP

8’116>50%

3 times cheaper than hybrid APCs��

~10 times cheaper for public purse� �

99.98% compliance

Publisher Journal APC

Nuclear Physics B $ 2’000

Physics Letters B $ 1’800

Advances in High Energy Physics $ 1’000

Chinese Physics C £ 1’000

Journal of Cosmology and Astroparticle Physics £ 1’400

New Journal of Physics £ 1’200

Acta Physica Polonica B € 500

Progress of Theoretical and Experimental Physics £ 1’000

European Physical Journal C € 1’500

Journal of High Energy Physics € 1’200

Average effective APC 2014:(In 2014 SCOAP3 pays max 2011 #articles, rest free)

€ 1’042

SCOAP3journals

APC(2014,inEuro)

4,000

3,500

3,000

2,500

2,000

1,500

1,000

500

Sources:JournalCita?onReport,publishers’websites,scoap3.org,webarchive.org

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18ImpactFactor(2012)

Chart:C.Romeuetal.(2014)TheSCOAP3ini1a1veandtheOpenAccess-Ar1cle-Processing-Chargemarket:globalpartnershipandcompe11onimprovevalueinthedissemina1onofscienceDOI:10.2314/CERN/C26P.W9DT

a)  hUps://github.com/OpenAPC/openapc-de;b)  hUp://figshare.com/ar?cles/2015_Jan_June_UK_APC_data_combined/1509860c)  hUp://blog.wellcome.ac.uk/2015/03/03/the-reckoning-an-analysis-of-wellcome-

trust-open-access-spend-2013-14/

AverageAPC2014paidbyGermanuniversi?es:€1,234a

SCOAP3averageeffec?veAPC2014:€1,042

AverageAPC2015paidbyUKhighereduca?oninst:€2,351b

AverageAPC2013-2014paidbytheWellcomeTrust:€2,502c

Realcoststothesystem(veryliUlefreshmoney)

Partnership Dec 2013 Partners joined during 2014 Partners joining in 2015

Partnership today: 43 countries + 3 IGOs Austria Belgium Canada CERNa

China Czech Republic Denmark Finland France Germany Greece Hong Kong Hungary IAEAb

NEW: Iceland Israel Italy Japan JINRc

Korea Mexico Netherlands Norway Poland Portugal Slovak Republic South Africa Spain Sweden Switzerland NEW: Taiwan Turkey United Kingdom United States of America NEW: 12 additional U.S. Universities

a) European Organization for Nuclear Research, Geneva b) International Atomic Energy Agency, Vienna c) Joint Institute for Nuclear Research, Dubna representing 12 of its member states

46 countries, 3 IGO, 3’000 libraries today, and growing

Research intensive countries supporting SCOAP3

Territory size shows the proportion of all scientific papers published in 2001 written by authors living there http://www.worldmapper.org/display.php?selected=205

SCOAP3Partner

OthercountrieswithatleastoneSCOAP3author

18’000 authors from 90 countries

Piled Higher and Deeper by Jorge Cham www.phdcomics.com

title: "Sharing" - originally published 9/4/2015

phdcomics.com/comics/archive.php?comicid=1818

Piled Higher and Deeper by Jorge Cham www.phdcomics.com

title: "Sharing" - originally published 9/4/2015

data

Open Data in High-Energy Physics �is impossible�

(too complex, large, dangerous, un-understandable, useless…)

Observer, Alamy

L3 detector at CERN LEP accelerator

hep-ex/0406049; 10.1016/j.physrep.2004.07.002

hep-ex/0406049; 10.1016/j.physrep.2004.07.002

hep-ex/0406049; 10.1016/j.physrep.2004.07.002

hep-ex/0406049; 10.1016/j.physrep.2004.07.002

hep-ex/0406049; 10.1016/j.physrep.2004.07.002

hep-ex/0406049; 10.1016/j.physrep.2004.07.002

hep-ex/0406049; 10.1016/j.physrep.2004.07.002

Welcome to CERN

To find the Higgs boson…

…analyse ~100PB(=100’000TB) on tape at CERN

What you really need to understand the Universe?

flickr.com/photos/mbiddulph/3240818979

Persistent Identifiers:no-profit, community-driven, infrastructures

5 Constraints on invisible width

Going beyond the e↵ective Higgs Lagrangian in Eq. 1, it is interesting to consider the possibilityof an invisible Higgs width. This may arise in models with new weakly interacting light degreesof freedom that have a significant couplings to the Higgs boson, for example in Higgs-portalmodels of dark matter or in supersymmetric models. The invisible decays have been directlysearched for at the LHC. The current 95% CL limits on the invisible branching fraction areBr

inv

< 65% in the ZH production mode in ATLAS 26, Brinv

< 75% in the ZH production modein CMS 27, and Br

inv

< 69% in the VBF production mode in CMS 28. Stronger limits on theinvisible Higgs width can be obtained indirectly from a global fit to the Higgs couplings. In thecase when the couplings of the Higgs to the SM matter take the SM values the invisible widthleads to a universal reduction of the decay rates in all the visible channels. This possibility isstrongly constrained, given the Higgs is observed in several channels with the rate close to theSM one. From Fig. 1(b) one can read o↵ the limit Br

inv

< 16% at 95% CL. This bound can berelaxed if one allows new physics to modify the Higgs couplings such that the Higgs productioncross-section is enhanced, so as to o↵set the reduction of the visible rates. For example, if cggis allowed to float freely in the fit, the weaker limit Br

inv

< 40% is obtained. Note that theseindirect limits apply to any other exotic (but not necessarily invisible) contribution to the Higgswidth.

Acknowledgments

AF thanks Dean Carmi, Erik Kuflik, Francesco Riva, Alfredo Urbano, Tomer Volansky andJure Zupan for collaboration on closely related projects. AF also thanks the organizers of theconference Windows on the Universe for the invitation and support.

References

1. G. Aad et al. [ATLAS Collaboration], Phys. Lett. B 716, 1 (2012) [arXiv:1207.7214[hep-ex]]. S. Chatrchyan et al. [CMS Collaboration], Phys. Lett. B 716, 30 (2012)[arXiv:1207.7235 [hep-ex]].

2. B. Grzadkowski et al. , JHEP 1010, 085 (2010) [arXiv:1008.4884 [hep-ph]],3. R. Contino et al. JHEP 1307 (2013) 035 [arXiv:1303.3876 [hep-ph]].4. A. Falkowski, F. Riva and A. Urbano, arXiv:1303.1812 [hep-ph].5. G. Aad et al. [ATLAS Collaboration], Phys. Lett. B 726 (2013) 88 [arXiv:1307.1427

[hep-ex]].6. ATLAS Collaboration, “Data from Figure 7 from: Measurements of Higgs boson produc-

tion and couplings in diboson final states with the ATLAS detector at the LHC: H ! ��,”http://doi.org/10.7484/INSPIREHEP.DATA.A78C.HK44

7. ATLAS Collaboration, “Data from Figure 7 from: Measurements of Higgs boson pro-duction and couplings in diboson final states with the ATLAS detector at the LHC:H ! ZZ⇤ ! 4`,” http://doi.org/10.7484/INSPIREHEP.DATA.RF5P.6M3K

8. ATLAS Collaboration, “Data from Figure 7 from: Measurements of Higgs boson pro-duction and couplings in diboson final states with the ATLAS detector at the LHC:H ! WW⇤ ! `⌫`⌫,” http://doi.org/10.7484/INSPIREHEP.DATA.26B4.TY5F

9. ATLAS Collaboration, ATLAS-CONF-2013-034.10. ATLAS Collaboration, ATLAS-CONF-2013-079.11. ATLAS Collaboration, ATLAS-CONF-2012-135.12. ATLAS Collaboration, ATLAS-CONF-2013-080.13. ATLAS Collaboration, ATLAS-CONF-2013-009.14. ATLAS Collaboration, ATLAS-CONF-2013-010.15. CMS Collaboration, CMS-HIG-13-001.

flickr.com/photos/nationalmuseumofamericanhistory/12619971753

Global community response

CMS detector

CMS collaboration: 3000 scientists, 85 countries - Consensus

CMS data preservation, re-use and open access policy

CMS data are unique and are the result of vast and long-term moral, human and financial investment by

the international community. There is unique scientific opportunity in re-using these data, at different

level of abstraction and at different points in time1. This opportunity calls for our collective responsibility,

and poses unprecedented challenges as no data sample of this complexity and value has ever been

preserved or made available for later re-use.

The CMS collaboration is committed to preserve its data, at different levels of complexity, and to allow

their re-use by a wide community including: collaboration members long after the data are taken,

experimental and theoretical HEP scientists who were not members of the collaboration, educational

and outreach initiatives, and citizen scientists in the general public.

CMS upholds the principle that open access to the data will, in the long term, allow the maximum

realization of their scientific potential. To that extent, CMS will provide open access to its data after a

suitable but relatively short embargo period, allowing CMS collaborators to fully exploit their scientific

potential.

This policy describes the CMS principles of data preservation, re-use and open access, as well as the

relevant actors in all these tasks and their roles and responsibilities. CMS understands that in order to

fully exploit all these re-use opportunities, immediate and continued resources are needed. The level of

support that CMS will be able to provide to external users depends on the available funding. This policy

addresses the moral responsibility of CMS for its data, as well as the increasing concern of funding

agencies worldwide and the civil society for the preservation and re-use of scientific data.

Notwithstanding the long-term perspective of the LHC programme, the time for action is now: lower-

energy and lower-luminosity LHC runs at centre-of-mass energies of 0.9, 2.36 and 7 TeV may never be

repeated, and their preservation and preparation for later re-use, has to be addressed urgently. Meeting

this challenge is a unique way to stress-test and evaluate the entire preservation, re-use and open access

concept for the CMS data.

CMS data take many forms. Starting from either raw experimental or simulated data through to

reconstructed data and the datasets of higher abstraction generated by analysis workflows, and finally all

the way to data represented in scientific publications. Each of these layers has the potential to afford

different opportunities for long-term re-use and poses different challenges for preservation. Data

represented in publications can already be preserved by building on the existing practices of the

Collaboration (e.g. open access publishing) and existing third-party platforms (e.g. INSPIRE2), simply

expanding the concept of publication to include additional data sets of a high level of abstraction. At the

other extreme of the spectrum, closer to the raw data, different challenges appear which imply a

paradigm shift from in-depth documenting and archiving of analyses during the publication process, to a

preservation of reconstruction and simulation software packages with all of their dependencies.

In general, four levels of complexity of HEP data have been identified, which map on to re-use

1A. Holzner, P. Igo-Kemenes, S. Mele, “First results from the PARSE.Insight project: HEP survey on data preservation, re-use and

(open) access” http://arxiv.org/abs/0906.04852http://inspirehep.net

DOI: 10.7483/OPENDATA.CMS.UDBF.JKR9

DOI: 10.7483/OPENDATA.CMS.UDBF.JKR9

G. Organtini

G. Organtini

G. Organtini

Wikipedia

flickr.com/photos/londonmatt/3163571645

flickr.com/photos/funky64/3925955346

Observer, Alamy

flickr.com/photos/hernanpc/16418578245

Salvatore.MELE@cern.ch