• Added 17 earthquakes between 1933 and 1949, all but one from the Gutenberg and Richter catalogue; these events are undetermined shocks in the ISS and with station data not in tabulated form; to relocate and reassess the magnitude of these events we digitized and integrated parametric station data from the ISS, Gutenberg notepads, individual station bulletins and BCIS; twelve of these earthquakes (evid 16963127, 16963144, 16963678, 16964661, 16964943, 16965068, 16965286, 16965324, 16966074, 16966380, 16966502, 16967106) are listed in the Main catalogue and are all shallow in the magnitude range 6.0 to 6.4 (many occurring in mid-oceanic ridges); the remaining five earthquakes have poor location and/or magnitude and are listed in the Supplementary catalogue (evid 16963156, 16964624, 16965441, 16965882, 16965899);
• The direct Mw and M₀ for evid = 16957859 and 16957865 (Tsetserleg and Bolnay 1905 great earthquakes in Mongolia, respectively) sourced from Schlupp and Cisternas (2007) are now correctly associated (in V4 the Mw and M₀ values were swapped). Thus the Tsetserleg (evid = 16957859) and Bolnay (evid = 16957865) earthquakes have Mw= 7.95±0.2 and earthquake 8.33±0.2, respectively. Also the seismic moment values have been updated accordingly. We thank Nobuo Hamada (formerly JMA) for pointing out the error;
• Improved location and magnitude for the 1905-11-08 earthquake (Aegean Sea, Greece, evid = 16957878). The epicentre moved about 122 km N-NE (near Mt Athos peninsula) from previous location, whereas the proxy Mw = 7.24 is now based on MS from five stations (instead of two in previous catalogue versions). The small uncertainty of the updated magnitude allowed us to move the event from the Supplementary to the Main Catalogue;
• Update location (about 270 km to the E) for evid 911802; also the magnitude has been improved since the Mw proxy is now based on MS from eight stations instead of five;
• Evid = 909805 is now listed in the Supplementary Catalogue as its location quality flag has been update to D; we thank Paolo Harabaglia from the University of Basilicata (Italy) for pointing out the large error ellipse semi-major axis for this event;
• Evid = 899720 now has a proxy Mw = 6.3 based on our re-computed MS after adding surface wave amplitudes from station HYD, India (MS from two stations, SVE and TIF, was not accepted in previous versions); the event is now listed in the Main Catalogue;
• Updated the seismic moment M₀ from the bibliographic search for evid = 828259 and 761994. In both cases the seismic moments from Richardson and Solomon (1977) were inserted with typos (1.12E+19 and 6.6E+18 Nm instead of 8.2E+18 and 6.4E+18 Nm for evid 828259 and 761194, respectively). Also the source reference for the seismic moment for evid = 828259 is now Singh et al. (1975) instead of Richardson and Solomon (1977). For both earthquakes the small variations in M₀ does not significantly change the moment magnitude Mw (for evid = 828259 Mw changes from 6.6 to 6.54);
• Replaced the Mw=6.5+/-0.2 from the bibliographic search for evid = 713088 with a proxy Mw=6.09+/-0.24 based on our re-computed mb;
• Minor changes due to the review done in the Rebuild project (Storchak et al., 2017):
  • For close events 684685 and 684686 we have moved the GCMT solution from 684685 to 684686; evid = 684685 is now listed in the Supplementary Catalogue as its Mw proxy based on our re-computed mb has high uncertainty;
  • replaced evid = 805431 (event previously in the Supplementary Catalogue) with evid = 805429 (listed in the Main Catalogue);
  • replaced evid = 856064 with evid = 856063.

Version 5.0 of the ISC-GEM Catalogue is the result of the 4th year of the Extension Project that began in November 2014. We included earthquakes smaller than 7.5 during 1904-1919 from a multitude of sources that were not processed in the past. We also included earthquakes that occurred in 2014. For earthquakes during 1904-1919 we had to digitize a multitude of station data sources from the ISA BAAS, ISS etc. and amplitude-period data added from original station bulletins to re-assess the magnitudes (mostly MS). The procedures used to digitize the parametric data is described in Di Giacomo et al. (2015). We did not add any station data before 1904 (basically only stations belonging to the Milne network are available, see, e.g., Adams, 1989) and, consequently, we decided to drop from the ISC-GEM catalogue the ten pre‑1904 events listed before Version 5 and have the catalogue starting in 1904.
Version 5.0 adds 521 earthquakes that occurred in 2014 with magnitude 5.5 and above and 1,110 for the period 1904-1919. For the latter period details of the extension are described in Di Giacomo et al. (2018).
We collected stations bulletins from various sources that complemented significantly the original ISC inventory of station bulletins. On top of the extension work, therefore, we decided to add amplitude data from the newly available stations bulletins also for the earthquakes that were already available in the ISC-GEM catalogue to improve the re‑computed MS and, as a result, the Mw proxy. With the data newly available, we better constrained the MS for 146 earthquakes between 1904 and 1919 listed in previous versions by using more station magnitudes (e.g., 80% of the original MS was obtained from NSTA ≤ 10, for the updated MS this percentage drops to ~40%). Finally, we revised the hypocentre locations and/or magnitude of 7 earthquakes outside the period 1904-1919. Here we list in detail the changes affecting these events:

• Evid = 16958009 has now a direct Mw = 7.1 following the paper by Pino et al. (2000);
• Evid = 16958130, also known as the 1911 Kemin earthquake has an updated direct Mw = 8.0 following the recent paper of Kulikova and Krüger (2015). Also, the location has been updated after adding several phase readings;
• Evid = 912062: updated location (~125 km to the SW) after considering the effects of the earthquake outlined in Ruiz and Madariaga (2018);
• Evid = 908523: update location (~40 toward NNW and depth changed from 15 to 65 km) after Paolo Harabaglia (University of Basilicata, Italy) pointed out the need of improving the hypocentre;
• Evid = 907513, also known as the 1930 Irpinia earthquake has now a direct Mw = 6.64 following the paper of Pino et al. (2008);
• Evid = 902493 has an updated location which fits better the tsunami generation; we thank Jonathan Griffin from Geoscience Australia for his enquire on the location of this event and following discussions;
• Evid = 722344, also known as the 1975 Hilo, Hawaii, earthquake has an updated direct Mw = 7.7 following the more recent paper of Nettles and Ekström (2004) instead of Kawakatsu (1989);
• Deleted event 650848 as part of the review done in the Rebuild project (Storchack et al., 2017);

We thank Paolo Harabaglia from the University of Basilicata (Italy) for pointing out the direct Mw in the papers for the two significant earthquakes in Italy and the 29 November 1975 Hilo (Hawaii) earthquake.

Version 4.0 of the ISC-GEM Catalogue is the result of the 3rd year of the Extension Project that began in November 2014. We included earthquakes smaller than 6.25 during 1920-1934 listed in the ISS but not processed in the past. We also included earthquakes that occurred in 2013. The recent data were available from the ISC Bulletin in an electronic format, but the arrival time data for earthquakes during 1920-1934 had to be digitized from the ISS and amplitude-period data added from original station bulletins in order to re-assess the magnitudes (mostly MS). The procedures used to digitize the parametric data is described in Di Giacomo et al. (2015).

Version 4.0 includes 484 earthquakes that occurred in 2013. For the period 1920-1934, we have relocated 3015 earthquakes and obtained new MS values for 2287 of them (available in the Main Catalogue). The events for which we could not re-compute the magnitude are listed in the Supplementary Catalogue.

As previously mentioned, we collected stations bulletins from various sources that complemented significantly the original ISC inventory of station bulletins. On top of the extension work, therefore, we decided to add amplitude data from the newly available stations bulletins also for the earthquakes that were already available in the ISC-GEM catalogue in order to improve our re‑computed MS and, as a result, the Mw proxy. With the data newly available, we better constrained the MS for 685 earthquakes between 1920 and 1934 listed in previous versions by using more station magnitudes (e.g., 67% of the original MS was obtained from NSTA ≤ 10, for the updated MS this percentage drops to ~20%). Finally, we revised the hypocentre locations and/or magnitude of 16 earthquakes outside the period 1920-1934. Here we list in detail the changes affecting these events:

• Evid = 16957859 and 16957865 (known also as Bolnai and Tsetserleg 1905 great earthquakes in Mongolia, respectively) have an updated direct Mw following the more recent paper by Schlupp and Cisternas (2007) instead of the Okal (1992) paper. Whilst Mw for evid = 16957859 changed only slightly from 8.3±0.4 to 8.33±0.2, Mw for evid = 16957865 changed from 8.3±0.4 to 7.95±0.2. The seismic moment values have been also updated accordingly;
• Banished event 805429 in the Supplementary Catalogue as it has been merged with evid = 805430;
• Magnitude re-assessed for close events 855990 and 855991. We kept the MS for evid = 855990 and dropped MS for evid 855991, which is now listed in the Supplementary Catalogue;
• Banished event 809906 as it has been merged to close event 809905;
• Magnitude re-assessed for close events 759121 and 759122. The new Mw=5.64±0.32 proxy for evid = 759121 is based on mb=5.53±0.17, whereas for evid = 759122 Mw=6.45±0.20 is still based on MS;
• Magnitude re-assessed for close events 759377 and 759378. The new Mw=5.62±0.54 proxy for evid = 759377 is based on mb=5.51±0.29, whereas for evid = 759377 Mw=6.15±0.20 is still based on MS;
• Magnitude re-assessed for close events 756556 and 756557. Readings contributing to MS for those two events have been re-associated and whilst the MW proxy is based on MS for both events, the new Mw proxy for evid = 756556 is highly uncertain (5.95±0.81) and mover evid = 756556 in the Supplementary catalogue, whereas the new Mw proxy for evid = 756557 is 5.90±0.20;
• Depth revision for three events in 2000: for evid 2348963 new depths is 55 km (previous was 36 km) whereas for evid 1912754 and 1741138 new depths is 100 km (previous depths were 25 and 85 km , respectively). The only change in magnitude is for evid = 1912754 where the Mw proxy is based on mb instead of MS and is highly uncertain (5.61 0.89). Thus evid = 1912754 is now listed in the Supplementary Catalogue.

Version 3.0 of the ISC-GEM Catalogue is the result of the 2nd year of the Extension Project that began in November 2014. We included earthquakes smaller than 6.25 during 1935-1949 listed in the ISS but not processed in the past. We also included earthquakes that occurred in 2012. The recent data were available from the ISC Bulletin in an electronic format, but the arrival time data for earthquakes during 1935-1949 had to be digitized from the ISS and amplitude-period data added from original station bulletins in order to re-assess the magnitudes (mostly MS). The procedures used to digitize the parametric data is described in Di Giacomo et al. (2015).

Version 3.0 includes 414 earthquakes that occurred in 2012. For the period 1935-1949, we have relocated 4403 earthquakes and obtained new MS values for 2719 of them (available in the Main Catalogue). The events for which we could not re-compute the magnitude are listed in the Supplementary Catalogue

During the year we also received station bulletins kindly provided by the personal collection of Prof. N. Ambraseys, British Geological Survey (BGS) collection, bulletins from the Geophysical Institute of the Czech Academy of Sciences in Prague, the Geophysical Survey of the Russian Academy of Science in Obninsk and the EOST of the University of Strasbourg (France). Those station bulletins helped to fill some significant gaps in the ISC original collection and allowed to add new surface wave amplitude-period data for earthquakes listed in previous versions of the ISC-GEM Catalogue. With the data newly available, we better constrained the MS for 648 earthquakes between 1935 and 1949 listed in previous versions by using more station magnitudes (e.g., 42% of the original MS was obtained from NSTA ≤ 5, for the updated MS this percentage drops to ~5%). Finally, we revised the hypocentre locations of 92 earthquakes outside the period 1935-1949. Most of the revisions concerns depth amendments from the first step (EHB location) of the two-tier location approach before the epicentre is relocated with the ISC locator (see Bondár et al., 2015). These 92 earthquakes have occurred between 1905 and 2004 but mostly before 1979; 34 of them were listed in the Supplementary Catalogue.

Removed evid = 877612 from the main catalogue as the event is a documented nuclear explosion in Novaya Zemlya.

Version 2.0 of the ISC-GEM Catalogue is the result of the 1st year of the Extension project that began in November 2013. We included earthquakes smaller than 6.25 from the 1950s, listed in the ISS but not processed by us in the past. We also included earthquakes that occurred in 2010 and 2011. The recent data were available from the ISC Bulletin in an electronic format, but the arrival time data for earthquakes in the 1950’s had to be digitized from the ISS and amplitude-period data added in order to re-assess the magnitudes (mostly MS), using procedures described in Di Giacomo et al. (2015).

Version 2.00, therefore, now includes data for 504 earthquakes from 2010 and 672 from 2011. For the 1950s, we processed and relocated 4156 earthquakes and obtained 2216 new MS and 12 mb values. We also introduced a new magnitude quality flag “D” for the Mw proxies with uncertainty larger than 0.7, which is now listed in the Supplementary Catalogue. In addition, since many ISS earthquakes during the 1950s when processed resulted in MS smaller than 5 (i.e, much smaller than our catalogue cut-off magnitude of 5.5), we have listed them in the Supplementary Catalogue rather than excluding them completely from the Catalogue, with particular regard to the importance of re-assessing the magnitude of earthquakes before the beginning of the digital era.

During the year we also received station bulletins kindly provided by colleagues from Edinburgh, Eskdalemuir, Obninsk, Prague and Apatity. Those station bulletins helped us to fill some significant gaps in the ISCs original collection and allowed us to add new amplitude-period data for earthquakes listed in previous versions of the ISC-GEM Catalogue. With the data newly available, we obtained fresh magnitudes for about 200 earthquakes that were previously listed in the Supplementary Catalogue (Version 1.05) and added 237 new earthquakes between 1960 and 1963. Finally, we revised the hypocentre locations of 458 earthquakes in the catalogue, predominantly in remote areas, with 75% of depth changes within 20km and the rest within 60km.

Updates to the catalogue regard 1) updated location and magnitude of the earthquake in Sumatra that occurred the 6 June 1943 (evid = 899872), which moved from the Supplementary Catalogue to the Main Catalogue; 2) updated the proxy MW for six earthquakes; 3a) addition of an earthquake (evid = 603952288) in the Main Catalogue that occurred in Turkey the 4 December 1905 and 3b) updated location of one of its aftershocks (evid = 16957879) that occurred approximately five hours later.

Details follow:

1) The location of the 6 June 1943 Sumatra earthquake has been updated and the epicentre moved to the North 40 km but most importantly the depth changed from 90 to a fixed depth of 10 km. Although the reported depth phase arrival times suggest that the hypocentre is around 90 km, we accept the 10 km depth since surface ruptures were reported (see, e.g., Untung et al., BSSA 1985). Hence, the event now moved from the Supplementary Catalogue to the Main Catalogue as with the shallow depth solution MS is available. From 12 stations we obtain MS= 7.8 +/- 0.1, which lead to a MW proxy of 7.81 +/- 0.25. We thank James Dewey (USGS Emeritus) for drawing our attention to this event.

2) the proxy MW for six earthquakes (evid 679878, 1656088, 1656090, 1728632, 13389548, 11911358) has been updated. With the exception of evid = 679878, the proxy MW is now based on mb instead of MS. The MS for those earthquakes was based on few stations with suspiciously large amplitudes. For evid = 679878 we obtained a new MS after correcting the surface wave amplitude units at six stations from micrometers to nanometers. The new six MW proxies are significantly smaller than the previous ones.

3a) the mainshock of the East Anatolian Fault sequence that occurred the 4 December 1905 has been added (evid = 603952288) after collecting arrival times from the ISA bulletin (61 arrivals) and surface wave amplitudes from four stations (GTT, LEI, POT, JEN). The event was missing due to a mis-association in the Centennial Catalogue between magnitude and origin time of the mainshock with one of its aftershocks;
3b) the location of the aftershock at 12h (evid = 16957879), which was already available in previous releases, has been updated. The previous ISC-GEM location was close to Aleppo, Syria, whereas the current location is much closer to the East Anatolian Fault.

We noticed that the Appendix is not recognised by our users as an important part of the ISC-GEM Catalogue. We therefore renamed the Appendix to be called Supplementary Catalogue.

The terms of the license have also changed to reflect the availability of the ISC-GEM Catalogue to all parties.

Updates to the catalogue regard 1) the proxy M_w for the 12 April, 1967 earthquake in the Leeward Islands region (evid = 836598) and 2) the GCMT solution for the complex event occurred near Potenza (South Italy) the 1990-05-05 (evid = 368231). With regard to the Leeward Islands 1967 event the original proxy M_w = 7.4 was based on the MS from three mis-associated station amplitude readings. The updated proxy M_w 5.0 is based on mb. We thank Uri Ten Brink from USGS office in Woods Hole, MA, for pointing out the excessive magnitude value. With regard to the complex Potenza earthquake of 1990-05-05, the GCMT solution has been moved to evid = 368231 from evid = 368230. These two events are separated by approximately 10 seconds and are interpreted as foreshock (evid = 368230) and mainshock (evid = 368231). For more details see, e.g., Ekström (1994). Since the magnitude estimates for the foreshock are uncertain, evid = 368230 is now listed in the Supplementary Catalogue. We thank Paolo Harabaglia (University of Basilicata, Potenza, Italy) for drawing our attention to these events and Göran Ekström (Lamont-Doherty Earth Observatory of Columbia University, USA) for discussing the association of the GCMT solution.

Updated the location for the earthquake occurred near coast of Venezuela the 1929-01-17 (evid = 907860) and the Mw for the 1967-07-30 Caracas earthquake (evid = 833882). The new location for the 1929-01-17 earthquake is about 98 km north the previous one and fits better with the tsunami observation for this event. The M_w = 7.0 for the 1967-07-30 Caracas earthquake initially adopted from Rial (JGR, 1978) has been replaced with a more recent determination of M _w = 6.6 by Suárez and Nábělek (JGR, 1990). We thank Michael Schmitz and Herbert Rendon from the Applied Geophysics Department, Fundación Venezolana de Investigaciones Sismológicas (FUNVISIS), for inquiring about the location and magnitude of these earthquakes.

Updated are the Mw proxy values for 84 earthquakes between 1988 and 2009. The M_w proxies were based on excessively large MS based on the surface wave amplitudes reported by just a few stations or a single agency. The updated M_w proxies are now based on mb. In particular, we thank Helene Lyon-Caen and colleagues from the Laboratoire de Géologie, Ecole Normale Supérieure in Paris, for pointing out the large M_w proxies for two earthquakes in Greece.

Updated the M_w and M_w proxy values of eight large earthquakes (evids 906508, 897799, 895348, 863756, 830410, 825732, 795586, 736783) following an additional review of corresponding scientific articles and correcting typographic errors in the results of the original bibliographical search for reliable seismic moments. In particular, we thank
Dr. Alexey Ivashchenko from the Tsunami Lab of the Institute of Oceanology, Russian Academy of Sciences, Moscow for pointing out an erroneous M_w for the Mondy earthquake (Baikal rift zone) of April 4, 1950 (evid = 895348).

This was the first public release of the Catalogue following the completion of the 27 months project.