In June 2020, the CMS collaboration submitted a paper titled “Observation of the production of three massive gauge bosons at √s= 13 TeV” to the arXiv preprint server. A scientific highlight in its own right, the paper also marked the collaboration’s thousandth publication. ATLAS is not far from reaching the same milestone, currently at 964 publications. With the rest of the LHC experiments taking the total number of papers to 2852, the first ten years of LHC operations have generated a bumper crop of new knowledge about the fundamental particles and interactions.
The publication landscape in high-energy physics (HEP) is very exceptional due to a long-held preprint culture. From the 1950s paper copies were kept in the well-known red cabinets outside the CERN Library (pictured), but since 1991 they have been stored electronically at arXiv.org. Preprint posting and actual journal publication tend to happen in parallel, and citations between all types of publications are compiled and counted in the INSPIRE system.
Particle physics has been at the forefront of the open-science movement, in publishing, software, hardware and, most recently, data. In 2004, former Director-General Robert Aymar encouraged the creation of SCOAP3 (Sponsoring Consortium for Open Access Publishing in Particle Physics) at CERN. Devoted to converting closed access HEP journals to open access, it has grown extensively and now has over 3000 libraries from 44 different countries. All original LHC research results have been published open access. The first collaboration articles by the four main experiments, describing the detector designs, and published in the Journal of Instrumentation, remain amongst the most cited articles from LHC collaborations and — despite being more than a decade old — are some of the most recently read articles of the journal.
Closer analysis
Since then, along with the 2852 publications by CERN’s LHC experiments, a further 380 papers have been written by individuals on behalf of the collaboration, and another 10,879 articles (preprints, conference proceedings, etc.) from the LHC experiments that were not published in a journal. However, this only represents part of the scientific relevance of the LHC. There were tens of thousands of papers published over the past decade that write about the LHC experiments, use their data or are based on the LHC findings. The papers published by the four experiments received on average 112 citations per paper, compared to an average of 41 citations per paper across all experimental papers indexed in INSPIRE and even 30 citations per paper across all HEP publications (4.8 million citations across 163,000 documents since 2008). Unsurprisingly, the number of citations peaks with the CMS and ATLAS papers on the Higgs discovery, with 10,910 and 11,195 citations respectively, which at the end of 2019 were the two most cited high-energy physics papers released in the past decade.
Large author numbers are another exceptional aspect of LHC-experiment publishing, with papers consistently carrying hundreds or even thousands of names. This culminated in a world record of 5,154 authors on a joint paper between CMS and ATLAS in 2015, which reduced the uncertainty on the measurement of the Higgs-boson mass to ±0.25%.
Teasing fluctuations
Ten years of LHC publications have established the Standard Model at unprecedented levels of precision. But they also reveal the hunger for new physics, as illustrated by the story of the 750 GeV diphoton ‘bump’. On 15 December 2015, ATLAS and CMS presented an anomaly in data that showed an excess of events at 750 GeV in proton collisions, fueling rumours a new particle could be showing itself. While the significance of the excess was only 2σ and 1.6σ respectively, theorists were quick to respond with an influx of hundreds of papers (see “750 shades of model building”). This excitement was however damped by the release of the August 2016 data, where there was no further sign of the anomaly, and it became commonly recognised as a statistical fluctuation – part and parcel of the scientific process, if ruining the fun for the theorists.
With the LHC to continue operations to the mid-2030s, and only around 6% of its expected total dataset collected so far, we can look forward to thousands more publications about nature’s basic constituents being placed in the public domain.
All numbers are correct of 7 January.
