Naghan (Chahar Mahal Bakhtiari-High Zagros, Iran) Earthquake of 6 april 1977. A preliminary field report and a seismotectonic discussion

The Naghan earthquake of magnitude 6 (Ms) occurred on 6 April 1977 in the mountanious area of Chahar Mahal Bakhtiari in the High Zagros, south of Shahr-e-Kord. It killed 348 people, injured about 200, and caused destruction over an area of 150 Km 2 . The shock damaged beyond repair 2,100 houses and killed 0.7°/o of livestock in the area; eight schools collapsed and 37 were damaged. The maximum intensity ot' the main shock did not exceed VIII (MM). The earthquake was associated neither with any fresh surface faulting, nor with reactivation of the existing faults and salt domes at surface. Heavy rain fell before and during the earthquake and the destruction was more extensive where the saturated clay content of the alluvium was higher; it was also more extensive in water-logged areas or the areas with a shallower water table. Landslides occurred on steep slopes and destroyed or damaged some villages built on slopes. Several aftershocks caused additional damage to the already destroyed or damaged villages. The Naghan earthquake was another instance

Folded Belt, indicating that the re-adjustment of the unexposed metamorphosed Precambrian Basement at depth caused no tectonic deformation (surface faulting) at the top of the sedimentary cover. This was due to the presence of the Upper Precambrian Hormoz layers (Salt deposits) acting as a slippage zone along the decollement surface of the Zagros at depth.

INTRODUCTION.
A strong earthquake of magnitude 6.5 hit the northern Zagros range (the High Zagros or thrust zone), heart of the Bakhtiari tribal area (Chahar Mahal Bakhtiari province), 100 kilometers southwest of Esfahan (Fig. 1) The instrumental location places the epicentre of the main shock 9 Km SSW of the macroseismc epicentre. The macroseismic epicentre (taken as the centre of the area of maximum destruction) is located at 31.85N, 50.76E. Several aftershocks of magnitude 4-5 followed the destructive shock and worsened the condition of the already damaged houses.
A preliminary examination of the area was made by a team of the Tectonic and Seismotectonic Research section of the Geological Survey of Iran, which concentrated on observation of the effects of the earthquake in built-up areas and the surrounding country.
Attention was particularly devoted to evidence of faulting or reactivation of faults and secondary effects such as cracks and openings in ground, lurching, tension cracks, temporary springs, and any other disturbances. Aerial photographs at a scale of 1 : 20,000 and 1 : 55,000, together with the Landsat (A and B) imageries and geological maps (B.P., 1963;IOOC, 1969) were used for locating the features observed and traces of recent and late Alpine Faults.
The intensity of the earthquake effect was estimated using the Modified Mercalli Intensity Scale. Conversation with inhabitants, their descriptions of the events, statistics, and information from the various authorities were used as guides.

Regional
Structure.
As previously stated, the region under consideration is situated in the High Zagros (Thrust Zone) of the Zagros Active Folded Belt. The geological history of this Belt is comparatively simple, as a relatively quiet sedimentation continued from Upper Precambrian (Infracambrian) to Pliocene. The Paleozoic sediments overlie a shield-like basement of Precambrian age. Regional disconformities occur at the top of the Aptian, the Cenomanian-Turonian, the Cretaceous, and the Eocene , James and Wynd 1965.
Evaporites have been prominently involved in Zagros tectonism. Jurassic and Miocene evaporitic sequences in the Zagros Active Folded Belt are zones above which disharmonic folding is reported, and similar movements have been postulated at the Hormoz Salt belts of Upper Precambrian age, near the base of the Paleozoic platform rocks, or on top of the metamorphosed Precambrian Basement (the metamorphosed Precambrian Basement does not crop out in Zagros). This important evaporite layer separates the sedimentary column from the basement. In addition, many large Hormoz Salt diapirs, some now active, reach the surface in the Zagros Active Folded Belt (Dunnington 1968a, b, Falcon 1974. The Zagros sediments covering the margin of the Arabian plate were folded mainly during the Main Zagros Orogeny (Wallachian diastrophism) at the end of Pliocene-Pleistocene times, when up to fourteen parallel anticlines and synclines were developed. These structures generally have a northeast-southeast direction, but in the extreme southeast (Fars-Hormozgan provinces) the trend of the fold axes gradually swings towards the east. The spectacular NW-SE striking folds (long, broad structures that are assymmetric or overturned towards the southwest, where they gradually die out) are cut by thrust and highangle reverse faults that are parallel to the folds but dip both northeast and southwest. These faults appear to be structurally related to the folds, and like them to be the results of plio-Pleistocene and Quaternary compression.
These deformations also generated along the northern side of the mountain range a major tectonic disconformity, the Main Zagros Reverse Fault, which separates the Zagros domain, in the southwest, from the Central Iranian domain.
According to Falcon (1974), the general picture of the Alpine Orongeny in the Zagros Belt is of a series of waves of depressions and uplifts advancing from the northeast towards the southwest. It is thought that the thrust Zone-Central Iran boundary was close to the northeast margin of the Arabian basement shield at least from the initiation of the Alpine movements.
From the crustal point of view the Simply Folded Belt contains simple folds caused by the rucking up above the decollement of the Phanerozoic column during one orogenic phase. The Thrust Zone is more complex because it has been rucked up twice: the first phase of folding and faulting affected the Thrust Zone in the Cenomanian Turonian to Middle Eocene interval, dominantly in Upper Cretaceous times. The Middle Eocene strata are of flyseh facies, while pre-Middle Miocene movements are relatively insignificant in the Simply Folded Belt. In the latter zone only a single phase of folding of Mio-Pliocene age has affected the rocks.
The very high twentieth century seismicity of the Zagros Active Folded Belt, is characterized by a large number of shocks in the magnitude 5 or 6 range, and a very small number with magnitude equal to or greater than 7. The precision of depth calculation of the Zagros is very poor, and without special studies little can be said on the subject, except that most earthquakes occur in the crust, and that they do not appear to exceed 100 km. Series of earthquakes resembling swarm also seem to be characteristic of the Zagros (Berberian 1976 a, b).
The total thickness of the sediments above the metamorphosed Precambrian basement cannot be more than 10 to 15 (or a possible maximum of 18) kilometers. Thus the earthquakes occurring within this depth range may be related to deformation of the sediments and or movements of the salt layers and their diapirism (?).
However, many earthquakes occur at greater depths, and salt diapirism would not necessarily be associated with thrust faulting, whereas the focal mechanism of most earthquakes are thrust faults. It would therefore appear that most of the earthquakes are subsedimentary and probably related to the basement or upper mantle (Haghipour et al., 1972) and that reactivation of some basement faults during readjustment of the basement causes subsedimentary earthquakes within the Zagros Active Folded Belt.

Previous Destructive Earthquakes of the Naghan Region.
In 1976, July 27 an earthquake of magnitude 5 was strongly felt at Shalamzar. The instrumental epicentre was located at 31.60N, 50.36E and the estimated focal depth was 77 km.
The only known semi destructive earthquake in Naghan region is the Sarpir earthquake of 21 September 1975. The instrumental epicentre of this shock, which took place at 14.16.37.8 GMT (17.46.36 local time), is located at 31.599N, 51.037E. The macroseismic epicentre was located at 31.228N, 51.04E (Fig. 2). The magnitude was 5.2 (Mb), the estimated focal depth 33 km (N) and the maximum intensity VI-VIII (MM). The earthquake destroyed Sarpir village and damaged Dorahan and Deh Bagh. Two people were killed in Sarpir.

EARTHQUAKE EFFECTS.
The following description of surface manifestations are represented in Fig. 2.

Faulting.
Study of the Landsat imageries (A and B) at a scale of 1: 1,000,000 and 1:500,000, aerial photographs of the region at a scale of 1 :20,000 and 1:55,000, the existing geological maps (B.P. 1963 I00C 1969) and detailed field study shows that there are only four majour faults in or near the Naghan epicentral region, namely the Ardal Fault, Shalamzar Fault, Dopolan Fault and Dena Fault.
Ardal Fault: This is a NW-SE reverse fault 80 km long, on the southeastern end of which the epicentral region is situated. It deepens towards the northeast. A linear Hormoz salt plug is intruded along its northwestern part. The fault is covered by Quaternary alluvial deposits.

Shalamzar Fault: This fault is situated north of the epicentral region. Its length is about 60 km and it deepens towards the northeast.
Dopolan Fault: This is another northwest-southeast fault 65 km southwest of the epicentral region. It starts from Dopolan in the NW and joins the Dena fault at its southeastern extremity.
Dena Fault: It is a NNW-SSE fault about 30 km southeast of the epicentral region. It seems that the macroseismic epicen- tre of Sarpir earthquake of 21.9.1975 was located on this fault. During a very careful investigation, these faults were checked in their entirety, and they were not reactivated at surface during the Naghan earthquake. Neither was the earthquake associated with the formation of new surface faults.

Salt Domes.
There are two late Precambrian salt domes (Hormoz salt) in the Naghan region intruded along Ardal and Dena faults: the Ardal salt dome in the north western part of the epicentral region, and the Dena salt dome in the southeastern part of the region (in Sarpir).
During the field investigation the contacts of the salt domes with the adjacent rocks have been carefully checked but no fresh movement has been found.

Landslides and Displacement of Stones.
Several large scale landslides occurred during the Naghan earthquake, including one about 1 km in length north and northwest of Ardal near the Ardal fault. The strike of this linear landslide is about N150° E.
Another landslide destroyed a vineyard and moved towards Sakiabad village, pushing, fissuring and destroying a few houses in the village (Fig. 3). Some landslides on the way to Ardal destroyed parts of the road.
Many stones in Quaternary alluvial deposits were upthrown and displaced during the earthquake (Fig. 4).

Rock falls.
Numerous rock falls occurred in different mountainous areas. Before and during the earthquake it was raining; consequently the rocks which fell during the earthquake left visible identations in the damp earth, clearly showing the path of the falling rocks.

Extensive destruction on saturated clay.
Due to the heavy rain before and during the earthquake, the clay alluvium of the epicentral region was saturated with water. It has been observed that the destruction was more extensive when the clay content of the alluvium was higher; it was also higher in water-logged areas or areas with a shallow water table. The northern part of Shalamzar village was not damaged but the southern part which has a very shallow water table (1 m), was damaged or partly destroyed.

Damage to houses and other
constructions.
The buildings in the Naghan region may be divided into the following groups: a) Abode-brick constructions with wooden beam roofs: These, together with the next group, comprise 90% of the houses in the area. The walls are of adobe-brick and clay mortar, and the roofs of wooden beams covered by clay and straw and tamped earth. New layers are added annually to the roof as a protection against rain and snow percolation. Most of these constructions are now completely destroyed, or are seriously damaged and unsafe as dwellings (Fig. 5). The high ratio of deaths to injuries emphasizes that the traditional adobe houses of the devastated region, with their weak walls and heavy earthen roofs, offer little resistance to strong earthquakes. Total collapse and heavy loss of life were normal in these buildings. The adobe houses were completely destroyed at Jaghdan (Fig. 6), while the stone masonry communal bath, with reinforced concrete ring beams was only slightly fissured (Fig. 7). b) Masonry constructions with wooden beam roofs: The mortar used here is often clay, so these constructions had a resistance similar to adobe-brick houses. The weakness of some of these buildings was further accentuated because each wall had been built virtually from two unanchored facings between which was a filling of earth (Fig. 8).
c) Masonry constructions with doomed roofs: In the area, there were a few structures of this type with thick walls supporting the arch roof. The usual mortar is gypsum plaster, or a mixture of gypsum and clay. d) Modern buildings with brick masonry and steel beam-Jack arch roofs: A few buildings of this type existed in the area. The school in Ardal was heavily damaged and partly destroyed (Figs. 9 e 10). SEISMOTECTONIC DISCUSSION. Berberian and Tchalenko (1975) and Berberian (1976 b, c) stated that there is no obvious relation between "Earthquakes" and "Structure" in the Zagros Active Folded Belt, and that it is not possible to use seismicity to establish smaller seismotectonic units with the information presently available. Neither the mountain/depression division, nor the large scale fault-flexure zones and strongly folded arcs, nor the mapped faults seem to exert any strong earthquake locations in the Zagros.
According to Berberian (1976 b) the apparent lack of agreement between earthquakes and structures may arise from two factors: "Deficiency of data" and "Unknown basement-surface relation".
The "seismological data" is defficient because of the brevity of the observation period (about 70 years) and the inaccuracy of focal depth determination (+ 30 km). The "geologicostructural data" is deficient because the tectonic history of fundamental structures, such as the Kazerun zone, had not yet been established by detailed field work. The major problem, however, is the unknown relation between the "Precambrian metamorphosed basement" at depth where most earthquakes presumably occur, and the "top of the sedimentary cover" whe-re the geological structures are observed. Between basement and cover, a "highly plastic salt layer (the Hormoz Salt)" acts as a "zone of slippage" on the decollement surface, disconnecting the cover structures from those at depth. Consequently surface structures may not always reflect the underlying basement structures in a simple manner.
Even within the sedimentary cover, other plastic layers, such as the Gachsaran Formation, are responsible for remarkable structural changes: synclines at surface correspond to anticlines 1 km down; faults change from high angle reverse to near horizontal thrust, or die out altogether, etc.
The Naghan earthquake of 6 April 1977 of magnitude 7 is another "Subsedimentary Zagros-Type Earthquake" case in the Zagros Active Folded Belt, showing that the reactivation of some basement faults at depth during the readjustment of the basement caused no tectonic deformation at the top of the sedimentary layers because of the presence of Hormoz plastic deposits. The thickness of the top sedimentary cover in the Naghan region is about six kilometers (Morris 1977). This six kilometers sedimentary cover overlies the Hormoz salt beds and together overlie the metamorphosed Precambrian basement.
As mentioned, the reactivation of the metamorphosed Precambrian basement caused no structural effect on the top of these six kilometers of sedimentary folded and faulted layers. No reactivation along faults or salt domes have been observed, and no fresh surface break was formed. The subsedimentary Zagros-type earthquakes which are characteristic of the Zagros Active Folded Belt, while having no relation to the top of the sedimentary cover, represent an important phenomenon which creates curious problems in the seismic zoning of this part of the country. It seems that the Zagros Active Folded Belt should neither be divided into several seismotectonic zones in the manner of Nowroozi (1976), nor given several expected maximum intensities for future earthquakes in different zones.
The Khurgu (North Bandar Abbas, Iran) earthquake of 21 March 1977 (Berberian, Papastamatiou, 1977) of magnitude 7.0, which occurred a few days before Naghan earthquake, was ano-ther "subsedimentary Zagros-type earthquake" which reinforces the present conclusions. These events showed that earthquakes may occur anywhere in the Zagros Active Folded Belt and since no correlation could be established between the seismicity and the tectonic features with our present knowledge, it should be assumed that an event has the same probability of occurrence anywhere within the Zagros Active Folded Belt. That is to say, the seismic risk level is the same everywhere along the Belt. Hence 20th century epicentre maps maybe dangerously unsuitable for the purpose of subdividing the Belt (and the country) into active and non-active regions. Reliably catalogued historical (pre-1900) earthquakes are so few as to be of little help. It would therefore be wrong to assume a very low risk rate in areas with no known record of a strong shock in the Zagros Active Folded Belt.