Historical Seismograms - Evidence from the AD 2000 Izu Islands ...

226 M. DeBecker and T. CamelbeeckAt the beginning of 1909, a 1000 kg Wiechert astatic pendulum, made by Spindlerand Hoyer of Gottingen, was added together with a clock that was changed dailyby telephone to the Royal Observatory time-service. A timing accuracy of about1 sec was obtained. The first vertical instrument, a 1300 kg Wiechert pendulum bythe same makers, was added in May 1910. The equipment was further augmentedthe following year by the addition of two classical Galitein horizontal instruments,made by Masing in St. Petersburg (Galitzin, 1911).In May 1912, the Bosch tronometers were given to the Ministry for the Coloniesand installed at Elisabethville, Congo (now Lubumbashi, Zaire).Instrumental constants were regularly determined, and are reported in the publishedmonthly bulletins, so that the seismograms retain their full use. The Wiechertsproved to be very sensitive to temperature changes in the vault, but the periodsof the Galitzins remained very stable, at just under 25 sec. In June 1913, the clockwas electrically synchronized with the Royal Observatory time-service, and timingaccuracy to 0.1 sec became possible.Two further additions to the instrumentation were made. In April 1930, a verticalGalitein-Wilip with a period of about 8 sec was installed. Its sensitivity was to beimproved and its period lengthened to 10 sec in 1936, when the original steel springswere replaced by Elinvar. Three Sprengnether instruments obtained in the early1950’s are still in use.In 1984, the ‘Centre de GQophysique Interne” began a systematic study of Belgianseismicity, which involved an inspection of all the available seismograms ofshocks before 1950, including those made abroad. Table 1 shows preliminary resultsfor the two best-documented earthquakes, a magnitude Ms = 5 shock in theBrabant massif on 11 June 1938, and the earthquake of 3 April 1949. The first ofthese was the largest in Belgium since 1800, and the second, a magnitude smaller,is representative of shocks in the Hainaut province, where it caused considerabledamage. The earthquake of 11 June 1938 is similar in magnitude to the biggesthistorical events in this region, as shown in Figure 1. The correct assessment of theseismic hazard in Belgium needs especially a detailed study of those main historicalevents. Concerning the 11 June 1938 earthquake, original seismograms or copieswere collected from 19 stations (Figure 2). Epicentres have been determined usingwell-determined P, arrival times, at 19 stations at distances from 140 to 800 kmin the first case, and 10 stations from 140 to 500 km in the second case (Table 2).A P,, velocity of 8.1 km/sec was adopted as an appropriate average value for theregion. Adopted focal depths were derived from macroseismic data as Uccle, thenearest station, was more than 50 km from the epicentre.Table 1. Characteristics of the Two Earthquakes StudiedDate Origin time Lat. Long. Depth M S mbLgON OE (km)1 1938 Jun 11 10h 57m 33s 50.78 3.59 15-25 5.32h0.17(10) 5.59 h0.19 (11)2 1949 Apr 03 12h 33m 41s 50.45 4.11 shallow 4.273Z0.31(05) 4.59 3Z0.25 (07)The numbers in parentheses indicate the number of stations used for the computationof these parameters.