CHAPTER 2

Arctic Seismological Data Bank

Arctic Seismological Data Bank (ARS) is part of a uniform bank system created in VNIIOkeangeologia for storing geoscientific information on the Arctic, Antarctic, and World¸s oceans. The above system is developed using personal computers which makes it available for every day use. A necessity in such system has been evident, but unfortunately, domestic technology has just recently allowed its implementation.

At present, over the entire period of instrumental observations, some 7 thousand earthquakes have been recorded north of the Arctic Circle alone, while in total there are 18 thousand including off-polar regions. The aforesaid suggests that this considerable set of data is scattered in the great number of foreign and domestic catalogues published on the basis of observational data obtained from global and regional seismographic networks. The catalogues are not unified by the form of the data presented, and they are based on different amount of reference data and different processing methods. Parameters of the same earthquakes in different publications usually differ, discrepancies in earthquakes of the first half of the century sometimes reaching hundreds kilometers. Besides, a great amount of factual seismological information, particularly abundant data on focal mechanisms, is scattered in a number of other foreign and domestic publications. For certain individual events, up to 5-6 determinations are known to be obtained by different authors who used different methods and frequently got essential discrepancies in the results. This scarcity of seismological information have made map-making procedure, even with respect to simple maps of epicenters, into a complicated and tiresome multistage process of information sampling, its evaluation in terms of precision and representation, correlation and matching of non-unified data, etc., which naturally has caused many mistakes. In addition, the precision of the compilation considerably downgrades when at its final stage, seismological information is manually applied on the cartographic base.

International Data Center (IDC) established in Moscow under the auspices of the Russian Academy of Sciences hosting seismological information as well, is definitely not in a position to solve all the problems for the following reasons:

- IDC is known to cover not all the Arctic earthquakes, which is particularly true for the Arctic regions outside Russia;

- IDC¸s information on earthquakes is incomplete. It is particularly true for energy characteristics and macroseismic data;

- IDC facilitating map-making process regarding epicenters is not efficient in the solution of other seismological problems such as the necessity to study statistic characteristics of seismic regimes or topical question of the unification of earthquakes energy characteristics;

- attachment to IDC like to any other central and monopoly office prevents flexible and prompt operations. It is evident that a normal research process may only take place in the event when the complete information is in hand or readily available.

The establishment and availability of a deliberately designed and thoroughly compiled data bank on the basis of modern PCs have largely solved the aforesaid problems.

2.1 Principal methods to fill the data bank

The ARS data bank was developed not only as a source of largely known information, that is virtually a PC-based catalogue, but also a system incorporating new information and permitting a number of computing operations to be performed in order to obtain characteristics required for the study of the Arctic seismicity.

A favorable factor in the development of seismological data banks is that the existing numerous catalogues of earthquakes are virtually ready-made forms for presentation of seismological information where the main scope of parameters is distinctly outlined to describe each earthquake.

At the same time, a specific feature of this information creating certain difficulties is in the aforesaid discordance since it comes from different seismological agencies. This situation provides the possibility to choose between two alternative approaches in the selection of seismological material when compiling data bank, on which the structure of the bank would depend.

The essence of the first approach is that any time when there are several solutions for one earthquake, only the most authentic one should be selected and placed in the data bank. At first sight, this approach seems to be rather attractive, since it makes the data bank compact and allows recurrence to be avoided, however, in practice, it has a number of disadvantages preventing its full employment. As a matter of fact, not always a preference of any solution is justified. It is especially true for the period of early 1960s, that is prior to the publication of annual official earthquake catalogues. This ambiguity in the evaluation has remained nowadays. Data supplied by powerful international seismological centers such as ISC and NEIC, on the one hand, are more reliable than those of regional agencies, due to both more advanced methods and processing technologies, and greater amount of initial material. On the other hand, they may be disadvantageous because of missing regional geological setting of every region, particularly, local time distance curves. Furthermore, in contrast to global, regional agencies frequently use materials of temporary stations located immediately in the epicentral zones and providing most valuable information. Sometimes, it happens that for the same earthquake different parameters are more representative according to different agencies or authors, or one source providing more dependable data on a part of the parameters, e.g. epicentral location, may not contain other parameters, such as magnitude or depth.

Another alternative approach is inclusion in the data bank of the entire available information. However, this approach despite its objective and unbiased nature, seems to be practically unsuitable. The reason is complication in use of such bank, and this is not a matter of drastic increase of its size. The main obstacle is in big difficulties arising from the development of criteria to select in every particular case one of many solutions regarding each earthquake for further operations with the bank, e.g. when compiling a map of epicenters

The optimum is the third way which was employed by ARS. It ensures, in our view, the maximum possible application of advantages of both above approaches. This allows the completeness and objectivity of the information to be combined with simplicity and reliability of its use. Its essence clearly understood at direct operation with the bank is as follows:

The main step is to divide the whole data set into units, each corresponding in principal to a certain information source, namely, the data of the global network (the general) are separated from the data of regional networks. The above suggests that the purest separation is only possible for the material obtained after the early 1960s when appropriate catalogues publication began with respect to each observational network. As a result, the division of the information into units have allowed the employment of the main advantage of the second approach, that is assured complete and unbiased information and elimination of its main shortcoming caused by difficulty in removing recurrences when the information is subsequently used. A certain increase of the machine memory consumption caused by the occurrence of the same earthquake in two units (the general and one of the regional) will be compensated by the possibility of its more efficient use. As it will be shown below, there are proper ways of determining earthquake magnitudes typical of each particular region and not applicable to any other. If information had not been divided into units, a space for all types of magnitudes should have been provided, and in many cases it would have been blank.

Prior to the early 1960s, when the division of the information by source was complicated and the number of nearly equal sources exceeded two, the approach employed for ARS establishment was close to the first one of the two approaches mentioned above. The most complete source was selected as of the time range to be recognized as the priority. The names of the priority sources are referred to in the seismological description of the data bank. In case of expediency of usage of a non-priority source for any particular event a special reference was made in the Notes. For instance, information on Canadian earthquakes for the period prior to 1957 was taken not from the catalogue by N.A.Linden (1959) which had been the priority for this time range, but from Canadian sources Employment of the above approach has permitted, in our view, complete and unbiased information to be provided without considerably complicating records and increasing the volume of the data bank.

In order to avoid drastic increase of the data bank volume, it was decided to only indicate errors of major parameter of the earthquake rather than each of them, namely to indicate absolute value (in km) of the epicentral location error.

For the purpose of data unification regarding focal mechanisms, computer determination and re-determination of 101 Arctic earthquake focus mechanisms (prior to 1991) was made using a single procedure (Aptekman and others,1979), that is ARS is also a source of new information (partially published in [Avetisov,1993a;1993b])

At present, in the ARS, the total number of earthquakes in the sampling, logarithm of this number and cumulative magnitude, that is parameters which provide the base for computation and compilation of magnitude-frequency, and for the investigation of seismic regimes, has become available for calculation. In addition, empirical formulas connecting different determinations of earthquake magnitudes may be computed as the basic prerequisite for the investigation of seismic regimes.

2.2. Data bank contents

ARS is divided into 5 databases (catalogues) : General, Arctic Canada, Northern Yakutia, Fennoscandia and Sevmorgeo, the number of which may be easily increased, if necessary. This seemingly illogical subdivision judging from the titles ensures, as it will be seen below, the most optimum data differentiation both by processing quality and by region.

The layout of the articles in each catalogue is unified to a maximum extent, while existing distinctions pertain to the information on energy characteristics of earthquakes and focal mechanisms.

Each catalogue contains the date (month, day, year), time in the focus (hours, minutes, seconds), epicentral localization accuracy, depth of the hypocenter (where known), magnitude or energy class (where available), the number of recording stations and the range of epicentral distances (where known), the number of positive C and negative (D) first motions signs ( for events recorded by at least 100 stations or supplied with focal mechanism solutions), macroseismic data and notes.

For any earthquake ARS presents all available determinations of magnitude and energy class, however the form of presentation varies. In each catalogues save for Sevmorgeo, special boxes have been provided for 5-6 magnitude modifications which are most frequently used in the catalogue. All other values are listed in the Notes.

Each regional catalogue presents all earthquakes of a relevant area regardless of the source of information, also including events involved in the General Catalogue (in this case a mark "AG" in the Notes stands - available in the General Catalogue).

The General Catalogue is the only to present information on available focal mechanism solutions. In other catalogues availability of focal mechanism solution is reported in the Notes.

The General Catalogue dates back to 1908. The main source of information up to the first half of year 1958 inclusive was the catalogue by N.A.Linden (1959). Apart from this, catalogues by J.H.Hodgson et al. (1965), L.R.Sykes (1965), New Catalogue ...(1977) were used. Where any of the above sources other than that by N.A.Linden is involved, an appropriate reference is given in the Notes.

Between the second half of year 1958 through 1963, the catalogue by L.R.Sykes (1965) had been the main source of information. Where other source is involved, an appropriate reference is provided.

Since 1964, the General Catalogue has been solely based on ISC (International Seismological Center) determinations (Regional Catalogue). Owing to this all information on earthquakes contained therein is unified subject to methods and ways of original data processing.

The General Catalogue was originally presumed to cover the region within the confines of the Arctic Circle. However, in the course of work it became clear that this geographical restriction would downgrade ARS since it would not make possible for the user to gain understanding of the relationships between seismically active zones of the Arctic and other regions on the Earth. In addition, this approach would make users when presenting cartographic materials, use but round-shaped charts, otherwise artificial zones without earthquakes would appear on the maps. Therefore ISC data were entered into the General Catalogue (i.e. beginning 1964) regarding the Central and Southern Alaska, northern Kamchatka and Sea of Okhotsk, and Komandorsky Islands. However, in view of high seismic activity of these regions seeking to avoid overloading of ARS with earthquakes of the Pacific seismic belt, only sufficiently strong events recorded by at least 25 stations had been selected from the above. As a result, the General Catalogue contains information obtained from the aforesaid sources before 1964, and since 1964 - the entire information obtained form ISC regarding earthquakes within the Arctic Circle and the most reliable date on more southerly events falling within the quadrangular plane table with angular points located as follows (latitude, longitude): 1. 53° ,-32 ° ; 2. 49° , -115° ; 3. 53° , 162° ; 4. 63° , 65° .

Geographic names are most often presented subject to the chart by E.A. Flinn and E.R. Engdahl (1965; 1974). An exception is made for "The Laptev Sea Area" incorporating the following regions included in the aforesaid chart : "Offshore Central Siberia", "East of Severnaya Zemlya", "The Laptev Sea", "Offshore northern East Siberia".

Special magnitude boxes indicate the most frequently occurring in ISC catalogues determinations made by ISC, USA National Earthquake Information Center and the Institute of the Earth Physics RAS in Moscow from compressive waves using short-period instruments, and surface waves using the vertical component of long-period instruments: mb(ISC), mb(NEIC), mb(MOS), Ms(ISC), Ms(NEIC), and Ms(MOS), respectively.

Parameters of focal mechanisms are presented in unified international format (Aptekman and others,1987) to which solutions published prior to its adoption are adjusted where possible. The author and year determined are also presented together with the type of applied materials or way of determination, type of dislocation, seismic moment (mainly for solutions obtained through centroid moment tensor CMT and moment tensor MT), and the quality of determinations (except for CMT and MT solutions), either taken from the authors or evaluated by the quantity and uniformity of distribution by quadrants, and by the number of inconsistent signs of first motions ( this information is also presented.) In the latter case, per cent ratio of consistent signs of first motions to its total number is computed and one of three group of reliability is selected for reference: I - > 89%, II - 80-89%, III - < 80%.

In either of the two cases when the total number of first motions signs does not exceed 40 either in one quadrant consistent signs are missing the group of reliability is lowered by one step down. In Comments auxiliary information dealing with the accuracy of focal mechanism solution is presented.

The Catalogue Arctic Canada includes earthquakes from 1912 within a hexagonal plane table with angular points located as follows (latitude, longitude): 1. 64° , -145° ; 2. 85° , -145° ; 3. 85° , -55° ; 4. 60° , -55° ; 5. 60° , - 130° ; 6. 64° , - 130° . Sources of information for the period up to 1962 inclusive catalogues by N.A.Linden (1959), J.H.Hodgson et al. (1965), L.R.Sykes (1965) had served. In each case appropriate reference is given in the Notes.

Since 1963 collections of works "Canadian earthquakes" over appropriate periods and ISC catalogues have been used as sources of information. The latter are based on data from regions 1, 2, 18, 19, 443, 444, 634, 635, 636, 675-682 (Flinn and Engdahl,1965; Flinn and others,1974).

For the overwhelming majority of Canadian earthquakes subject to their location, epicentral distance and wave period length, the Seismological Observatory in Ottawa determined magnitudes from one of the following regional scales: O.Nuttli(1974) for maximum Lg-wave at the vertical component of short-period instruments Mn(OTT) and K.Richter (Gutenberg and Richter.1956) for recording S and Lg-waves by similar instruments ML(OTT.) For Alaska earthquakes, ML magnitude determined by the Alaska Tsunami Warning Center ML(PMR) is presented. Six special boxes accommodate these three magnitude modifications as well as mb(ISC), mb(NEIC), and Ms(NEIC).

The Catalogue Northern Yakutia includes earthquakes beginning 1914. Sources of information up to 1963 inclusive had been catalogues by N.A.Linden (1959), New catalogue (1977), J.H.Hodgson et al. (1965), L.R.Sykes (1965). In each case corresponding reference is given in the Notes.

Since 1964 collections of works "Earthquakes in the USSR" over appropriate periods and ISC catalogues have been used as sources of information. The latter are based on data from regions 654, 655, 667-670 (Flinn and Engdahl,1965; Flinn and others,1974).

The catalogue includes earthquakes within the quadrangular plane table with angular points located at (latitude, longitude): 1. 65° , 105° ; 2. 75° , 105° ; 3. 75° , 150° ; 4. 65° , 150° .

For most earthquakes in the catalogue, only K class value has been determined due to which it occupies one of the five special boxes for energy characteristics . The remaining four boxes are intended for mb(ISC), mb(MOS), Ms(ISC), and Ms(MOS.)

The Catalogue Fennoscandia includes earthquakes beginning 1911. Sources of information up to 1950 inclusive had been catalogues by N.A.Linden (1959), J.H.Hodgson et al. (1965), L.R.Sykes (1965), New catalogue (1977), M.Bath (1956). In each case corresponding reference is given in the Notes.

Since 1951 catalogues by G.D.Panasenko "Earthquakes in Fennoscandia" ( up to 1985 inclusive) (Panasenko,1977; 1979; 1986; 1991), collections of works "Earthquakes in the USSR" and ISC catalogues have been used as sources of information. The latter are based on data from regions 534-537, 542, 642, 646, 721-724.

The catalogue includes earthquakes within the heptagonal plane table with angular points located at (latitude, longitude): 1. 62° , 0° ; 2. 55° , 4° ; 3. 55° , 18° ; 4. 60° , 30° ; 5. 66° , 45° ; 6. 72° , 36° ; 7. 72° , 10° .

Earthquakes in Fennoscandia are recorded by a number of seismic agencies under various jurisdiction each presenting their own energy characteristics of earthquakes. Contribution of one or another agency to this business has changed in time. It is quite evident that the question of magnitude unification in this region is the most topical. In the catalogue Fennoscandia apart from mb(ISC) and MS (ISC) six special boxes for energy characteristics are occupied by determinations obtained from the Bergen Seismological Observatory based on the duration of the record Md(BER) and local scale ML(BER), the Uppsala Seismological Institute based on the local scale ML(UPP) and data from G.D.Panasenko on the recording range Md (APA.)

The Notes refers to the agency where information has been obtained from, if it is not provided by catalogues "Earthquakes in Fennoscandia" for the period of 1951 to 1985 inclusive or determinations by ISC after 1985 (e.g. "From UPP data").

The Notes also indicate the code of the station, if parameters of the earthquakes were determined from data obtained by one station.

The catalogue Sevmorgeologia basically includes only earthquakes recorded by NIIGA-SEVMORGEO field stations (See Appendix.) Most of them were not recorded by any station in the Global or a regional network. Apart from the aforesaid earthquakes, the catalogue contains a few weak earthquakes recorded by the Kheis station on the Franz Joseph Land.

No boxes have been provided for the magnitude. Information available on energy characteristics is presented in the Notes.

The Notes also present the way of epicenter location determination .

As it has been mentioned above, data on the Arctic seismic stations and seismic agencies have been entered in ARS as auxiliary information. ISC catalogues were used an a source of information for these databases.

The data bank contains both effective and closed national and foreign stations now or previously located north of 60° N. For each station, the name, international code, geographic location and height above the sea level in meters are presented.

The Notes indicate to which jurisdiction the station belongs, date (year) founded and effective period; from time to time auxiliary information is provided on the involvement of the stations in the regional networks, instrumental features, etc.

The data bank also contains all seismological centers ever mentioned in ARS. For each seismological center abbreviation, name and mailing address are indicated.