1 Introduction

Search for kimberlite bodies is carried out by a variety of geological and geophysical methods, where ground and airborne magnetic surveys have the primary importance among all geophysical methods. Magnetic survey in simple conditions, where kimberlite hosting rocks crop out or are covered with thin (20 m) sediments, allows us coniidently identify magnetic anomalies associated with kimberlite pipes. The effectiveness of magnetic survey is reduced in areas covered with thick sediments or trap complex rocks. Also there are non-magnetic diamond-bearing pipes which can not be delineated by the magnetic surveys have been found, for example, the Nakyn field kimberlite bodies in the Yakut diamond province^[1].

The application of a package of geophysical methods is necessary to solve search problems in difficult geological conditions, as well as to detect weakly magnetic kimberlite pipes. They include electromagnetic exploration and development of techniques aimed at recognizing the anomalies due to pipes and selecting the structural features associated with the localization of kimberlites.

In previous years, investigations by various, mainly transient electromagnetic methods in depth study of the cross-section to 200~300 m, were carried out in exploration areas of the Yakut and Archangelsk diamond provinces^[2]. There was a considerable amount of field data however the electromagnetic methods efficiency in solving search problems is not high enough. Kimberlite pipes are low-contrast objects comparing to the hosting rocks. Overlying horizon s characterized by significant heterogeneity and the presence of local bodies of low resistivity, which create pipe type anomalies (tuff pockets in dolerite sills and lenses of clays and siltstones in terrigenous sediments). These search sites features put forward higher requirements to ensure the necessary depth, measurement accuracy and data interpretation reliability at presence of environment heterogeneity of the applied electromagnetic methods. Tensor methods application and high depth of investigations provide promising solutions to search problems. These approaches reduce the influence of shallow heterogeneities and improve the reliable identification of anomalies due to kimberlites.

Audiomagnetotelluric (AMT) sounding method based on measurements of natural electromagnetic fields in the frequency range from a few hertz to several thousands hertz, allows us to study the cross-section in the diamond areas in the depth range from 10's of meters to 2~3 km. In the AMT method the tensor measurements are implemented which allow us to obtain reliable results in horizontally heterogeneous media. Modern AMT equipment provides the high accuracy (2 % for apparent resistivity and 0.5 degrees for impedance phase), which makes it possible to identify the low-contrast objects in the cross-section. Using a simple model of the primary field (plane wave) provides reliable data interpretation.

The first testing on estimation of AMT method possibilities with the ACF-4 equipment for kimberlites exploration has been accomplished at objects of the Yakut and Archangelsk diamond provinces in 2000^[3]. Similar investigations have been carried out also using the equipment of Phoenix-Geophysics Ltd. ^[4].

In the past few years St. Petersburg State University has carried out investigations to assess the AMT method possibilities with the ACF-4M equipment for kimberlite exploration in areas of Yakut and Archangelsk diamond provinces. These works were aimed at the structure mapping tasks solving (identification of kimberlite fields, clusters of pipes and mapping of kimberlite control and hosting faults) and the identification of kimberlite bodies. The results are discussed in the article.

2 Geological and Physical Characteristics of the Research Objects 2.1 Yakut diamond province

The Yakut diamond province is situated in north-eastern part of the Siberian platform (Fig. 1). Its geological structure can be divided into three structural units corresponding to the main stages of the region development: the crystalline basement, kimberlite hosting carbonate rocks and the overlying terrigenous sediments. The crystalline basement of Precambrian age in the province central part according to the structural driling s at up to three or four kilometers depth. It is composed of Archean granites, granite-gneiss complex and the Archean and Lower Proterozoic crystalline gneisses and slates.

Kimberlite hosting formation is represented by carbonate and terrigenous-carbonate rocks of the Vend, Cambrian, Ordovician and Silurian ages with thicknesses varying from 1.5 to 2 km. The rocks lie almost horizontally. The structure of these rocks is complicated by a number of faults, which are a reflection of the crystalline basement block structure.

Overlying terrigenous sediments are composed of sandstones, siltstones and clays with interlayers and lenses of gravellits of Carboniferous, Permian and Jurassic ages and in some areas include different amounts of tuffaceous material. Thickness of overlying sediments varies from several meters to 100~150 m, in some cases up to 200~250 m. The characteristic feature of the overlaying layer is the presence of the trap complex rocks within it, which is a significant interfering factor in search operations.

Magmatic associations of the Yakut province are mainly groups of basic and alkaline-ultra basic rocks. The rocks of the trap complex are most widely spread and belong to the group of basic rocks. Kimberlite magmatic association has the second place after the trap, belonging to the group of alkaline-ultra basic rocks. Kimberlite bodies of the Yakut province are mainly of middle Paleozoic age (Devonian-Carboniferous) and the conditions of formation are divided into intrusive and explosive phase groups. Explosive facies bodies represent the explosion pipes and are most widely spread. The bodies of intrusive phase occupy a subordinate position and occur as dykes, sills and stocks.

Kimberlite pipes in the plan are isometric or oval in shape with axes ratio usually from 1:1 to 1:3 and the size from tens to several hundred meters. In the vertical section the pipe structures are divided into two parts: a funnel-shaped crater to a depth of 100~200 m and a channel close to a cylindrical form, gradually pinching-out, and at a depth of several hundred meters (sometimes 1~2 km) changing into the vein body. The crater contacts inclination angles range from 50° to 75°, the average cylindrical channel inclination angle is 82°~84° Funnel-shaped craters of some pipes in Yakut province were cut by erosion. Kimberlite dykes and veins are often found near pipes.

Kimberlite rocks are divided into three groups: 1) kimberlites or porphyric kimberlites, 2) eruptive kimberlite breccias, 3) kimberlite tuffs and tuff breccias. The vertical zonality is usual for the kimberlite pipes. In this case, their upper parts are composed of tuff breccias kimberlites, eruptive (autolithic) breccias are typical for the medium horizons, and massive porphyric kimberlites-for the lower ones (Fig. 2). Parts of the pipes formed by one kimberlite type can be layered, column forming or irregular in shape, but marked vertical zonality is the most characteristic feature for the pipes (Fig. 3). The small size pipes are usually formed by one kimberlite generation. Large pipes are typically multiphase. An essential part of kimberlite breccias is xenoliths of the host material and the deeper fragmented rocks. At times these xenoliths breccias can reach 80 % of the total volume of the pipe. The xenoliths content decreases in the deeper horizons, and increases in the near-contact zones, with the xenoliths size increasing.

Kimberlite fields are located at zones of intersecting faults that control kimberlite emplacement. Pipes within the kimberlite fields are usually grouped into the pipe clusters (usually from 2 to 10 bodies). However, they are placed along certain directions, forming linear groups (chains) of bodies. The long pipes axes are oriented in the same directions. In some cases there is a displacement of individual pipes at some distance (hundreds of meters few kilometers) from the direction along which most of the bodies of a cluster are located (Fig. 4a). The pipes usually have one root, but bodies with two or three roots are also found (Figs. 4b, 4c).

The impact of kimberlites on the host rocks is mainly mechanical, and thermal and chemical changes are of secondary importance. Cracks in the vicinity of kimberlite bodies are divided into primary that existed prior to their penetration, and secondary, accompanying the pipes formation. Primary faults served as facilitating channels for the kimberlite magma. The long pipes axes are usually oriented along the highest order faults, and the pipes are placed at the intersection of these faults with fractures of the lower orders. Fracturing accompanying the pipes formation includes a cracks system, located sub parallel and sub perpendicular to their contacts. These cracks are often filled with small kimberlite dykes. The subsidence structures in the pipes vicinity occur because of the increased rocks fracturing in the media near pipes and the subsequent karst processes. The amplitude of the vertical displacement of the hosting rock layers for the structures of subsidence in some cases can reach 50~100 m, and their sizes in plan are estimated as 2~4 times the pipe radius.

According to the bedding conditions and form among the traps the layered intrusive phase bodies dominate (dolerite sills) with a thickness of 30~60 m, and in some casesup to 120~170 m, as well as stocks and dykes. Layered bodies of effusive phase (covers) occur less frequently. Traps are younger formations in comparison to kimberlite pipes, and have the Upper Paleozoic -Lower Mesozoic age. There is a dependence on the dolerite sills shapes on the lithologic composition and bedding characteristics of the surrounding rocks. When intrusion the sills occur at the contact of pipes hosting carbonate rocks and overlying terrigenous sedimentary rocks they have layered form with thickness changes small in large areas (Fig. 5a). If the sills intrude into the overlying sediments, their form is more complicated, there are bulges, nicks and sharp thinning (Fig. 5b). The sills intruding into tuff rocks form a complex etwork, consisting of small size layers, interleaving and branching dykes.

2.2 Archangelsk diamond province

There are three structural units in the geological structure of the Archangelsk diamond province: Archaean-Lower Proterozoic age basement, presented by crystalline rocks, platform cover of Riphean-Paleozoic age with the thickness of 0.5~3.5 km, represented by alternating argillites, siltstones and sandstones hosting kimberlite pipes, and the overlying sand-clayey sediments with interbedded limestone up to 100 m of Carboniferous, Permian and Quaternary ages.

The kimberlite magmatism age in the Archangelsk province s estimated as Late Devonian-Early Carbonferous. In addition to the kimberlite pipes in the region pipes and dikes of basaltic composition close in age to the kmberlites are also found^[6].

By the spatial distribution of magmatic objects in the Archangelsk kimberlite province the kimberlite fields are selected (Fig. 6). The most studied is the Zolotitskoe kimberlite field in which the pipes are confined to the deep fault zone with the sub meridian strike and localized in areas of is intersection with faults of the north-west and sublatitudinal orientation. The directions of the long axes of pipes in a plan coincide with the deep faults.

Pipes penetrate the argillite-siltstone-sand strata. Their impact on the surrounding rocks is expressed as intense fracturing of the host medium near pipes in separate zones up to 50 m. In the host medium some of the layers rise with the amplitude up to 20 m to a distance of 40~50 m from the contact. In the upper part of the hosting rocks troughs of subsidence isometric shape with size several times larger than the diameter of the pipe are identified^[6]. Pipes of the Zolotitskoe field covered by unconsolidated terrigenous-sedimentary rocks of Quaternary age have the thickness from several to 70 m.

Most of the pipes are characterized by complex internal structure, depending on the formation conditions (single-phase, two-phase), on the crater presence or absence and on the crater phase rocks thickness. The pipes are usually isometric in shape, but some are also elongated and dumbbell-shaped. In their vertical section pipes have a conical shape, while some of them have a crater in the upper part. Pipe craters are usually filed with tuffaceous sedimentary rocks of the crater phase.

2.3 Electric properties of rocks

kimberlites are characterized by lower resistivity (ρ) values than the surrounding and overlying rocks, however, the electrical properties difere-nces are small and in some cases m inor. The rocks ρ values in the Yakut province mostly cover the following ranges: kimberlites -10~400 Ωm, dolerites -1000~20000 Ωm, tuffs -15~100 Ωm, the host rocks -150~7000 Ωm, overlying rocks -20~400 Ωm. The Archangelsk province is characterized by the following ρ values: kimberlites -20~60 Ωm, thehostrocks -100~200 Ωm, overlying rocks -100~300 Ωm.

The electric properties of the rocks show that the kimberlite pipes are low-contrast objects in relation to the surrounding rocks. Overlying rocks are characterized by a considerable heterogeneity level and the presence of low resistivity local bodies creating pipe type anomalies.

Different types of kimberlites are characterized by different resistivity values. The most conductive are porphyric kimberlites. Autolithic kimberlite breccias and tuff breccias are characterized by relatively high resistivity values.

Induction logging data (Fig. 7) showed increased electrical conductivity (σ) values of about 100 mS/m (ρ is about 10 Ωm) for the deepest part of the Komsomolskaya pipe folded by porphyric kimberlites, and low a values of about 20 mS/m (ρ is about 50 Ωm) for autolithic kmberlite breccias in the upper part. The magnetic susceptibility (χ) logging diagram illustrates the fact that higher χ values of kimberlites distort the a diagram on local intervals, but in general along the section the σ values can be estimated reliably. Distortions occur because of incomplete phase separation of the signal in induction logging equipment. It leads to negative values appearance in a diagram.

3 Hardware and Software Audiomag-netotelluric Soundings Complex

The audiomagnetotelluric soundings method -AMT is a high-frequency modification of the method of magnetotelluric sounding-MT^[7] and is based on natural electromagnetic fields measurements in the audio frequency range from a few hertz to several thousands hertz^[8]. The main sources of these fields are distant thunderstorms. According to the horizontal and orthogonal components of electric and magnetic fields measurements in the AMT method the amplitude of the surface impedance (or the apparent resistivity ρ_a) and impedance phase φ_z are determined. Using the apparent resistivity and the impedance phase values at different frequencies the sounding curves are obtained and as a result of inversion the geoelectric section is derived. The model of primary field in the form of a plane vertically incident wave s used for the AMT data interpretation.

In carrying out the work by AMT method in diamond bearing regions the ACF-4M system, developed by the St. Petersburg State University and MicroKOR Ltd., was used^[9]. The equipment includes a digital recorder with four synchronous channels and 24-bit ADC for each channel. The frequency range of equipment is 0.1~800 Hz, sampling frequencies are 160, 1600 and 3200 Hz, the volume of internal memory is 1024 MB. In the ACF-4M installation measuring parameters are performed using the station keypad or external computer. Field observations at the soundings stations are done by data recording into the internal memory. The time series and spectrogram are recorded; the apparent resistivity and impedance phase are computed directly in the recorder. Software control of the equipment provides visualization of the signals spectral characteristics produced on the recorder's display, and data quality evaluation s done directly on the sounding station. Work is carried out using a GPS receiver linked to the coordinates and time.

Hardware and software of ACF-4M system allows both scalar and tensor measurements. The scalar mode is typically used for rapid surveys in areas with horizontally-layered sections or when the orientation direction of the main geological structures is already known. In carrying out such work one electric line and one magnetic antenna are used. In the tensor mode measurement signals of two orthogonal electric lines and two magnetic antennas are performed. From results of tensor measurements we can get impedance tensor components and also the apparent resistivity and impedance phase curves both in the receiving measuring set directions and in the principal ones of the impedance tensor. Directions of AMT curves are determined by orientation of electric field receiving lines. Using the ACF-4M instrument t is possible to realize fast surveys (20~30 soundings per day depending on conditions of measurements-relief, etc.) with a field crew of 3 persons.

The software includes hardware control and data preprocessing SM27 program. This program produces a set of measurement parameters, the measured characteristics (spectrograms, the pair coherences) visualization on the data logger display, programmed control measurements, data storage and exports them to an external computer. Time-series processing by the SM27 program on the external computer includes the power auto spectra measured field components E_x, E_y, H_x, H_y and cross spectra based on Fast Fourier Transform (FFT) calculation. In the same program the apparent resistivity and the impedance phase curves are calculated by auto spectra and cross spectra, corresponding to two orthogonal azimuths of electric receiving lines. The SM27 program is used to process field data obtained at a low noise level.

To process the data obtained in difficult conditions (adverse noise conditions, a low level and stability of the natural electromagnetic field), the SM+ program is used. This program uses various types of robust procedures. As a result of data processing by this program there are the apparent resistivity and impedance phase curves on the receiving electric lines azimuth or principal directions of the impedance tensor.

For input data visualization and interpretation the Geoinf program is used. The program is designed for AMT data viewing and analysis on individual sounding stations, or profile and area and coordinate plan with the measurement stations image, pseudo-section of any parameters on the profile, the apparent resistivity and impedance phase curves for the H-and E-polarization field for a given set of frequencies along the profile, the distribution on the coordinate plan of any parameter on the area for a given frequency, geoelectric sections of profiles, the distribution of the inversion results (the values of resistivity) in area at a given depth plotting.

The method of effective linearization (MEL)^[10] and the MEL program for medium model with smooth changing properties with depth are used for 1D inversion. For 2D inversion the Shell2D program is used. The 2D inversion is based on the integral equations method by modified algorithm OCCAM^[11~13]. The program Shell2D allows the inversion for E-or H-polarization curves, or curves of both polarizations. The quality of the inversion was calculated by the generalized discrepancy between field and calculated curves. In favorable conditions with good quality field data and the proximity of the actual structure to two-dimensions, the misfit between the field and the calculated apparent resistivity is 1 %~2 %, and impedance phase difference is no more than 0.5~1 degrees.

4 Deliniation of Kimberlite Fields and Pipe Clusters

A reliable identification of the boundary of a kimberlite field is useful in launching a well focused search in the most promising areas. To solve this problem the studies to determine the complex geological, tectonic, geochemical and geophysical characteristics of kimberlite fields have been done. In this case, while using the geophysical data (seismic, gravity, magnetic and magnetotelluric sounding) the main attention is focused to study the deep stratifications in the Earth crust and mantle^[14~16]. According to geophysical data the pipe clusters are usually characterized by the same anomalies as kimberlite fields, but smaller in size and contrast.

Valuable additional information on the structural features within the kimberlite field and a pipe cluster may be obtained from a study of the structure of sedimentary cover from the surface to the crystalline basement. In the sedimentary cover kimberlite control and kimberlite hosting faults, characteristic domed structure for kimberlite fields, flexural bends and structure of subsidence occur. The structures of deeper layers in the crust and mantle which maybe indicative for kimberlite fields, clusters and pipes, are reflected in the sedimentary cover. Many researchers have pointed out that the sedimentary cover within the kimberlite field is characterized by increased fracturing associated with the presence of many differently oriented faults. The indications of kimberlite fields and pipe clusters presence in sedimentary cover can be identified and investigated more reliably and economically as compared to the direct investiga-ttons of these structures at great depths.

To identify the kimberlite fields and pipe clusters allocation the AMT method can be used to study the sedimentary cover from shallow depths (50~100 m) to the basement (2000~3000 m). The most informative is the study of regional distribution of reference horizons. The reference horizons, being conductive, can be reliably identified using AMT data. For example, in the Yakut diamond province the water saturated layer below the base of frozen rocks is used as reference. This horizon s usually in the depth interval from 200~400 to 500~800 m. The layer resistivity is reduced around 1 Ωm due to water highly mineralized and it can be confidently mapped by the AMT method. Bends, thickness abrupt changes and faults in this layer very well reflect the structural features of the sedimentary cover. In the Archangelsk diamond province in the AMT data contrastingly reflect conductive horizon of sandstones, siltstones and argillite, saturated with mineralized water, which occurs in the depth interval 500~900 m.

Layers of high resistivity can also be used as reference horizons. They do not make good targets as compared to conductive layers, however, due to a very high sensitivity of the AMT method to small changes in the degree of fractures in rocks filled with mineralized water, the study of these layers may provide valuable information about the sedimentary cover structure within the kimberlite field and pipe cluster.

4.1 Alaki-Markha field

To delineate the features in the sedimentary cover reflecting the presence of kimberlite fields and pipe clusters, AMT surveys were carried out along a north-western oriented regional profile in the Alakit-Markha kimberlite field of the Yakut province. The contour of the field according to the previous geological and geophysical studies is shown by the dotted line in Fig. 8. Within the Alakit-Markha kimberlite field about 60 pipes are discovered, including several of the industrial value. The pipes are grouped into a few clusters which are elongated in a north-eastern direction. Fig. 8 shows the clusters position and some of the larger pipes within them. Geological structures within the Alakit-Markha field (zone of faults, grabens and flexural bends are associated with the vertical block movements) have the north-eastern orientation.

The profile length was 74 km (619 stations of soundings). The distance between the sounding stations was 100~200 m, while in some detailed profile fragments 50 m. To determine the kimberlite field features the results of 2D inversion were used, which allow us to reduce the surface heterogeneities influence and to obtain more reliable data on the deep horizons. Analysis of the results of tensor measurements and data on the strike of geological structures show that the H-polarized field corresponds to the profile direction, and E-polarized field -across the regional profile.

For the inversion the amplitude and phase AMT curves for the H-polarized field in the direction of the regional profile (azimuth 325°) were used. As a result of the inversion the geoelectric section was derived (Fig. 9a). In Fig. 9b the section for a limited range resistivity variation (logρ=0~1.5, or ρ=1~32 Ωm) is shown for better visualization of the conductive horizon morphology. It should be noted that the geoelectric section shown in Fig. 9 is expanded along the vertical axis in a ratio 1:4. It must be kept in mind while evaluating the real morphology of the isolated layers. The figure also shows the position and the north-western border of the Alaki-Markha field according to earlier investigations.

The focus of the results analysis was given to tracking the reference horizon of highly mineralized waters under frozen rocks. This horizon has a regional distribution. It is characterized by reduced values of resistivity (2~10 Ωm) and contrastingly reflected in the AMT data.

In Figs. 9a, 9b the dotted line shows the position of the frozen rocks base assigned according to the hydrogeological drilling wells in the area. The position of water saturated layer determined from AMT survey coincides with the drilling data. In the frozen layer composition the conductive horizon of relatively small thickness (up to 100 m) is delineated. Induction logging data from the Alakit-Markha kimberlite field show that the resistivity of frozen rocks is determined by their lithologic content, and the delineated conductive horizon is made by marls. The distribution and thickness of frozen rocks in this area is not a determining factor affecting the rocks electrical properties. The resistivity values depend mainly on their water saturation and lithology.

From this section it is obvious that within the kimberlite field the thickness of the water-saturated layer varies from 300 to 450 m (from the absolute mark +150 to -300 m). This water-saturated layer is characterized by considerable variability in thickness and in some intervals we see a reduction of is thickness up to 200 m. Outside the kimberlite field a decrease of water-saturated layer thickness up to 250 m (from the absolute mark +200 to -50 m) is observed, but the water-saturated layer thickness is stable along the strike. Such features as thickness, morphology and occurrence features of the reference horizon highly mineralized waters are determined by the tectonic fragmentation in the rocks, which is higher within the kimberlite field. They are contrastingly reflected in the AMT data and can be used to locate the kimberlite field boundaries.

The regional profile site, where the cluster of the Yubileinaya kimberlite pipe s located, appears contrastingly on the geoelectric section in the form of a local vertical anomaly of resistivity low values. It should be noted that this anomalous zone has a length of 5 km, it is mapped very clearly (50 stations of soundings) and is not influenced by local surface heterogeneity. Other clusters of kimberlite bodies are identified by the AMT data less confidently. This is explained by the fact that only the Yubileinaya kimberlite pipe cluster directly crossed by the regional profile and the other clusters are at some distance from it.

4.2 Zolotitskoe field

The Zolotitskoe field in the Archangelsk diamond province includes several kimberlite pipes, located in a chain along the meridian direction (Fig. 6), which can be considered as pipes cluster. As a result of AMT data 2D inversion for H-polarized field (latitudinal direction) the geoelectric section was derived along the latitudinal profile crossing the kimberlite hosting zone in the region of the Koltsovskaya pipe (Fig. 10).

Earlier seismic surveys delineated the presence of flexural bend in the basement roof (shown by ashed line in Fig. 9), which contains the kimberlite hosting fault and the pipes cluster of the Zolotitskoe kimberlite field. Flexural bends in the basement are usually accompanied by fracturing zones and are favorable areas for the kimberlite magama intrusion. As seen from the figure, the presence of flexural bend also appears as the logρ isolines bend at the geoelectric section according to AMT data.

The most contrast feature of the pipe cluster is the behavior of conductive horizon in the depth interval 500~900 m, associated with mineralized water-saturated sandstones, siltstones and argillites, which in this case can be considered as reference. The conductive horizon rising by about 100~150 m is observed at the western and eastern sides at 500~700 m distance from the kimberlite hosting fault. This increase s apparently linked to the kimberlite hosting fractured zone filled with mineralized water.

4.3 Mirny field

In the western part of the Mirny field in the Yakut diamond province the studies by AMT method were carried out for the possibility estimation of locating the cluster, which composed of five pipes including the kimberlite bodies Tayozhnaya, AN-21 and Amakinskaya. These kimberlite bodies are localized in a chain in the meridian direction along the Zapadny kimberlite controlling fault (Fig. 11a). The depth of the basement in this area is about 2000 m.

The most informative parameter that reflects the overall site structure s the impedance phase. As it is seen from Fig. 11a kimberlite bodies are located along the eastern conductive zone boundary marked by increased values of the effective phase. The boundaries of the conductive zone in the northern and central parts of the site coincide with the kimberlite controlling fault Zapadny but turns in a westerly direction in the southern part.

Fig. 11b shows the geoelectric section according to the 2D inversion for the H-polarized field along the profile 30, crossing the area in the latitudinal direction. As can be seen from the figure, the considered conductive zone corresponds to an increase in thickness of the conductive horizon saturated with under frozen water in the depth interval 300~600 m, and a decrease in resistivity of the underlying salt-bearing rocks in the depth interval 600~1400 m. According to the logρ isolines behavior at depths of 1800~2200 m a flexural bend s observed, which coincides with the Zapadny kimberlite controlling fault zone.

The resistivity decrease in the high-resistivity salt rocks is associated with the increased fragmentation of the sedimentary cover and fling the cracks with conductive saline water. The presence of faults within the conductive zone shows a change in the morphology of under frozen conductive horizon, abrupt changes in is thickness and depth. As the reference in this case, the high-resistivity salt rocks horizon can be considered. It has the stable enough values of resistivity outside the conductive zone, which controls the eastern border of pipe cluster and the low resistivity within the conductive zone.

4.4 Nakyn field

The Nakyn kimberlite field is located within a system of deep faults with the north-eastern direction. According to seismic investigations it corresponds to the flex shape zone which is an area of crystalline basement sinking from the north-west to the south-east from 2.5 to 5.0 km. In the kimberlite pipes localization area the basement depth is 3.7~4.0 km.

Four kimberlite bodies are located in the Nakyn field along the north-eastern strike hidden kimberlite hosting fault Diagonalny which is directed at an angle to kimberlite controlling faults (Fig. 12). Its mapping is based on a detailed study of kimberlite hosting rocks cores by the presence of a number of micro breccias, micro faults, tectonic cleavage, mirrors sliding^[17]. The location of kimberlite bodies and of their long axes orientation emphasizes the presence of this fault. The fault Diagonalny is also characterized by low-contrast anomalies according to the results of the detailed magnetic and shallow seismic surveys. The Nakyn kimberlite field can be considered as a pipe cluster. The Nyurbinskaya and Botuobinskaya are pipes, and the Markhinskaya and Maiskaya are vein bodies (shown in Fig. 12 with different views).

Investigations by the AMT method were performed on the zone of the Markhinskaya and Maiskaya kimberlite bodies with the 200 m distance between the profiles and 60 m -between the sounding stations (Fig. 13). It is interesting to consider the AMT data features in the zone with the width of about 500 m (shown by dotted lines) which contains kimberlite bodies. The overall structure of the site is well reflected in the impedance phase isolines scheme for H-polarized field (Fig. 13a), connected with the features of the area deep (at a depth of 2~3 km) structure. At the same time, kimberlite bodies located along the conductive zone border which is distinguished by high values of the impedance phase in the site north-western part. The conductive zone coincides with the flexural bend in the basement roof and is associated with increased fracturing of sedimentary cover rocks. The localization features of Markhinskaya and Maiskaya kimberlite bodies in the Nakyn field are the same as discussed above for the cluster of pipes Tayozhnaya, AN-21 and Amakinskaya in the Mirny field.

The transition zone from low to high impedance phase values is oriented at an angle to the zone which contains the Markhinskaya and Maiskaya kimberlite bodies. This is probably due to the shingling structure of the Diagonalny kimberlite hosting fault (Fig. 12). The fault segment containing the Nyurbinskaya, Botuobinskaya and Markhinskaya bodies is more clearly manifested in the impedance phase solines scheme.

The location zone of Markhinskaya and Maiskaya bodies is observed in the resistivity isolines scheme according to the 1D inversion for the H-polarized field in a relatively shallow depth of 200 m (Fig. 13b). Within this zone linear conductive areas located at an angle to the zone direction are observed, and they are apparently associated with cleavage cracks formed as a result of horizontal movements of rocks blocks.

5 Faults Mapping

Considering the structural control of the kimberlites location most researchers associate kimberlite fields with zones of deep faults. They are identified by the faults series of similar strike and most often filled with steeply-dipping trap dikes. This type of faults controlling kimberlite placement is confidently mapped by magnetic surveys.

The pipes within the kimberlite fields are usually located along certain directions, forming linear groups (chains) of bodies. The pipes long axes are often oriented in the same directions. These pipes location features indicate the presence of another type faults -kimberlite hosting ones. Their detection by geophysical methods s more difficult. In these faults there are no trap dykes, and they are not distinguished by strong anomalies of the magnetic field. So the role of electromagnetic methods for mapping kimberlite hosting faults s increasing.

Fig. 14 shows the geoelectric section to a depth of 600 m according to the AMT latitudinal curves (H-polarization) 1D-inversion results along the latitudinal profile through the Koltsovskaya kimberlite pipe in the Zolotitskoe field (the Archangelsk province). Tectonic faults previously identified on seismic data (shown by vertical solid lines) are located by local conductive anomalies. Good coincidence or proximity of faults identified by the previous seismic investigation results with anomalous geoelectric section sites allows us to identify these faults quite well according to AMT data.

As can be seen from Fig. 14 the AMT curves 1D inversion results can not distinguish the kimberlite hosting fault in which the Koltsovskaya pipe is located from other faults. Abnormal appearance of the kimberlite hosting fault is the same as other faults on this profile. As shown above, it is more informative when location of the kimberlite hosting fault on this profile is done employing 2D inversion of AMT data and tracing the morphology of sandstones, siltstones and argillites deep water-saturated horizon.

Experience with the application of AMT method in the Yakut diamond province showed that tracking of the reference under frozen horizon, saturated with conductive mineralized water, is the most informative for the faults allocation. These faults coincide with sites of breaks, sharp changes in thickness or inclination of the water-saturated horizon.

6 Mapping of Kimberlite Pipes

Kimberlite pipes generally occur in three different types of geological settings in Russia. The first group includes the exposed pipes (generally cropping out to the surface or covered with a thin, up to 20 m, sediments layer), the second-covered with sediments of medium (20~50 m) and high thickness (50~200 m), the third -covered with rocks of the trap complex.

6.1 Pipes at the open areas

For exploration of kimberlites of the first group with electromagnetic methods, including the AMT one, usually do not cause difficulties. The following examples illustrate the characteristics of pipes anomalies allocation in AMT data without interference of overlying sediments and rocks of the trap complex.

An example of the confident delineation of the pipe that crops out to the surface s the results obtained on the Geokhimicheskaya pipe, the Alakit-Markha field in the Yakut diamond province (Fig. 15). According to the prospecting studies it has previously been established that the Geokhimicheskaya pipe cropping out to the surface is slightly elongated in a north-eastern direction form in its plan. Its size in the plan is 130 × 180 m. Drilling results up to 105 m show that the pipe is composed of the autolithic kimberlite breccias.

AMT studies at the Geokhimicheskaya pipe were performed on six profiles of north-western direction and with 900 m length and with the separation between sounding profiles and stations -100 m. Near the pipe the distance between sounding stations was 50 m. The data inversion for ID medium model was done by the MEL program. The results of AMT surveys and the mathematical modeling showed that the kimberlite pipes anomalies mapped by assuming 1D model are more informative as compared to those based on 2D model.

The profiles, sounding stations and logρ_a and φ_z isolines schemes for E-polarized field (the minimum apparent resistivity curves) at the 130 Hz frequency are shown in Figs. 15a, 15b. It follows from these figures that the kimberlite pipe as a conductive object is easily mapped through low logρ_a and high φ_z values. It may be noted that the pipe contour at the surface and logρ_a and φ_z anomalies are shifted. It is an indication of the inclined bedding of the pipe and is incidence in the south-eastern direction. The noticeable shift of impedance phase anomaly, which is linked to deeper parts of the section than the apparent resistivity, is a confirmation of the inclined disposition of the pipe.

Fig. 15c shows geoelectric sections for logρ along the 1 and 2 profiles according to the 1D inversion for E-polarized field (the minimum apparent resistivity curves, mainly oriented on 70° azimuth). The Geokh imcheskaya pipe can be seen in the area marked by reduced values of resistivity. The upper part of the pipe, up to a depth 100 m, is characterizedby the ρ values 40~100 Ωm (logρ=16~2). The more conductive pipe's part is in depth range from 100 m to 250 m, where the ρ=10~40 Ωm (logρ=1~1.6). The most conductive and deepest part of the pipe is the depth range from 250 to 400~600 m with low values of ρ in the range of 2.5~10 Ωm (logρ=0.4~1). It seems that the deepest conductive part of the pipe is composed of porphyric Itimberlites.

At the geoelectric section along the profile 1 at the station 14 a deep conductive anomaly (at a depth of about 400 m) is identified. According to drilling it wasfound that the low values of the resistivity are associated here with the fracture zone.

The Akademicheskaya and Ilmenitovaya pipes (the Daldyn field in the Yakut diamond province) also crop out to the surface. Fig. 16a shows the pipes position and AMT profiles. The geoelectric sections obtained according to the 1D inversion of AMT data for H-polarized fields (the maximum apparent resistivity curves, mainly oriented on 45° azimuth). Local anomalies of low resistivity values are seen within and outside the boundaries of pipes from the geological data (Figs. 16b, 16c). The features of pipes anomalies on the sections are their non uniform internal structure for the resistivity, which is due, prob-ably, to different types of kimberlites forming the pipes and the presence of sharp heterogeneities in the near-contact zones of pipes. Some of the conductive anomalies at this site are connected with fractured rocks in the fault zones.

6.2 Pipes covered with sediments

An example of locating the kimberlite pipes covered with sediments is illustrated by the results of AMT survey in the Zolotitskoe field in the Archangelsk diamond province. Fig. 17a shows the pipes location in the area and AMT stations. The pipes are found at a depth of about 50 m. Figs. 17b~17d show the geoelectric sections obtained from the 1D inversion of AMT data.

In the sections the kimberlite pipes appear as conductive anomalies. The most conductive kimberlite blocks occur at great depths and are associated with the root parts. In some cases there are conductive anomaly displacements at the contacts of pipes (the Pionerskaya pipe, Fig. 17b). This is probably due to the presence of various kimberlite types within the pipes. The conductive porphyric kimberlite composes often the pipes root parts.

In 1D inversion sometimes the influence of 3D heterogeneities is appeared. It leads to higher apparent depth of pipes location. Mathematical modelling reveals that the higher apparent depth is typical for 1D inversion results, and the apparent depth of the upper edge of a pipe body may be enhanced by 50~100 m.

Near the western contact of the Pionerskaya pipe (Fig. 17b) a heterogeneous zone is observed, which, as noted above, is the typical feature of pipe delineation by the AMT data.

Anomalies of pipes usually appear in the AMT data while using both H-and E-polarized field. As seen from Figs. 17d, 17e the Pervomayskaya pipe is characterized by contrasting anomalies for H-and E-polarized field. In practice, H-polarization is the most frequently used parameter, because it maps too the tectonic faults more reliably.

One of the challenges that arise when searching for the pipes is to distinguish between the anomalies due to pipes and tectonic faults. The task is complicated by the fact that the pipes are usually associated with faults showing good electric contrast according to AMT results. As seen from Fig. 14, anomaly location of the Koltsovskaya pipe is slightly different from the anomalies associated with adjacent faults. For reliable separation of anomalies due to pipes and tectonic faults it is necessary to carry out the area survey. At the same time, it is possible to recommend the exploration drilling of similar anomalies because of their quantity s not so great and they are very promising for new kimberlite pipes discovering.

Tensor measurements and the use of two polarizations can bring significant benefit for solving this problem. Fig. 18 shows that the Koltsovskaya pipe displays good contrast for both polarizations. At the same time, the anomalies associated with faults can be located in the geoelectric section at one polarization and not in the other. The fault near station 7 is determined with the contrasting conductive anomaly in the H-polarized field (Fig. 18a) and is not detected for the case of E-polarization (Fig. 18b). Similarly around station -14 the fault yields a good anomaly in the H-and in the E-polarized field, which is probably due to the influence of faults intersection zone according to the seismic data.

The different features of the Koltsovskaya pipe anomalies for the cases of H-and E-polarization are due to the fact that the pipe s elongated in shape and the profile intersects it at an angle (Fig. 18a). The pipe side effect s more noticeable for the E-polarization.

6.3 Pipes covered with traps

Kimberlite pipes covered with rocks of the trap complex are the most difficult objects for electromagnetic exploration methods. The rocks of the trap intrusive phase-dolerites are characterized by resistivity high values (2000~20000 Ωm). Explosive phase rocks -tuffs, and sediments, occurring as pockets and lenses in dolerite sills, have low resistivity values, mostly15~50 Ωm. The great resistivity contrast of in relation to the dolerite and the location of the pockets and lenses of tuff and sediments close to the surface lead to their significant influence on the electromagnetic methods results, including to AMT ones, and hinder the kimberlite body identification.

Difficult conditions for the AMT method are areas covered with traps and sediments. As noted above, trap sills in this case are characterized by great variability in thickness, presence of squeezes and windows (Fig. 5b).

When dolerites overly the kimberlite hosting rocks the sills have simple form and stable thickness (Fig. 5a). Such conditions are more favorable for the application of AMT method. However, the presence of tuff and sediments pockets and lenses in dolerite sills complicate the investigations. At same time, when the volume of tuffs and sedimentary rocks is small the conditions for applying the AMT method are favorable.

It should be noted that for the search sites on which the dolerites overly the kimberlite hosting rocks the application of geophysical methods to explore for kimberlites assumes importance. Under these conditions the mineralogical methods of kimberlites exploration are not useful because of the lack of overlying sediments and the weathering zones of carbonate hosting rocks which contain minerals associated with diamonds.

Small size pipes are usually single-phase and composed of kimberlite breccias with increased values of resistivity. Their identification according AMT data at sites overlapped by traps is unlikely.

Large pipes are usually multiphase. They are composed of both kimberlite breccias and porphyric kimberlites with lower resistivity values. In favorable conditions (when dolerites cover the kimberlite hosting rocks, small volume of tuffs and sedimentary rocks in the sills), the pipe can be successfully identified according to AMT data.

The Schukin pipe (the Alakit-Markha field, the Yakut province) is covered with rocks of the trap complex with the thickness of about 30~40 m and terrigenous sediments with the thickness of about 50~70 m (Fig. 19a). Apart from the main sill, the low-thickness (about 1 m) dolerite apophysis is also presented in the section near the contact of the hosting carbonate rocks and overlying sediments which crosses the pipe upper part.

According to the drilling data the Schukin pipe has two phases containing porphyric kimberlites (north-western part of the pipe) and autolithic kimberlite breccias (south-eastern part of the pipe). It should be noted that the vertical section of the pipe (Fig. 19a) and plan (Fig. 19b) are very approximate, as plotted according to the small number of wells (only three of them crossed the pipe). Real form of the pipe in the section and plan could differ significantly from those shown in Fig. 19a.

Fig. 19 shows the resistivity isoline scheme at a 200 m depth (Fig. 19b) and geoelectric sections for some profiles (Figs. 19c~19f) obtained from 1D inversion results for the H-polarized field (azimuth 325°). On geoelectric sections pipe appears on the profiles 70.5, 71, 71.5. In this case, the characteristic conductive anomalies are partially located within the north-western part of the pipe selected according to the drilling, composed by porphyric kimberlites, and partly shifted outside the contour in a north-western direction. South-eastern part of the pipe, composed by autolithic kimberlite breccias, not detected on the crossing it 71. 5 profile.

In Fig. 19b we can observe two anomalous zones-one within the north-western part of the pipe, folded by porphyric kimberlites, and the second-outside, shifted in a north-western direction. These data show that the real form of pipe differs from the contour shown in Fig. 19b, and perhaps there is another root, shifted in a north-western direction. Similar to Fig. 4a, for this pipes luster too, including the Schukin pipe, the multiroot structure of the pipes could laterally shift some roots within the main kimberlite hosting zone.

However, in the area of the Schukin pipe the conductive anomaly of the kimberlite hosting fault is allocated, which is superimposed on the pipe anomaly and makes its selection difficult. This anomaly gradually increases from the north-east to south-west due to the presence of rocks with high degree of fracture and water saturation in the fault zone. To interpret AMT data in such conditions it is necessary to choose the most informative parameters that highlight the anomaly due to the pipe, such as the polarization, the level of cut-off depth up to which there is a good anomaly of kimberlites in relation to the surrounding rocks, etc. In this example, the cut-off depth of 200 m is optimal since at greater depths the influence of the heterogeneity of the upper part of the section, and the contrast of kimberlites and water saturated under frozen conductive rocks is negligible.

AMT investigations were performed on site of the Zarya pipe, which is covered with trap complex rocks of about 100 m. Traps cover the kimberlite hosting carbonate rocks. In the dolerite sill are pockets of tuff and lenses of sediments. According to the drilling the Zarya pipe is multiphase. The central part of the pipe body composed with autolithic kimberlite breccias, and the near-contact zones-with porphyric kimberlites.

Fig. 20a shows a scheme of AMT profiles and stations and the projected outline of the pipe according to the drilling. Geoelectric sections are obtained from the results of 1D inversion using the data for E-polarization (the minimum apparent resistivity curves, mainly according to 55° azimuth). At the geoelectric section (Fig. 20b), corresponding to the pipe part delineated by drilling, the high conductivity anomaly locates part of the pipe composed of porphyric kimberlites. The central part of the pipe body, composed of autolithic kimberlite breccias, has a higher value of the resistivity which is not reflected in the AMT data (Fig. 20d, profile 00). At the section along the profile 00 two anomalous conductive zones are seen (north-eastern and south-western edges of the pipe).Form of the anomalous zones indicates inclined bedding of conductive kimberlite blocks.

In the profile-1 the characteristic pipe type anomaly is observed outside the contours of the pipe, shown on the plan and section from earlier data. It is possible to assume the presence of an additional root of the pipe. Multi-root structure of the kimberlite bodies is a characteristic feature of the pipe cluster. Located in close proximity to the pipe Zarya, the kimberlite pipes Podtrappovaya and Aikhai have respectively two and three roots (Figs. 4b, 4c).

7 Conclusions

Application of the AMT method provides new possibilities for the exploration and structural mapping tasks solution in diamond regions. The AMT investigation of the sedimentary cover section from 30~50 m to the basement (2~3 km) shows the important features of the kimberlite hosting rocks, pipe clusters and tectonic faults both controlling and hosting kimberlite pipes. Using this method the anomalies prospective to locate the kimberlite bodies in the difficult geological conditions are established. The AMT method allows us to obtain really valuable information about the hosting rocks structure and features, which can not be achieved by the transient electromagnetic methods with shallow (about 200~300 m) investigation capabilities.

The location of the kimberlite fields and pipe clusters is based upon the reference layers that are mainly conductive and are delineated well by the AMT method, or the layers of the high resistivity. That show insignificant variations of the fissuring and fluid saturation degree in the geoelectric sections. Outside the region of kimberlite fields, the reference horizons have the stable thickness and horizontal bedding. Within the fields these horizons are characterized by the increased thickness and its variations reflect increased fracturing of kimberlite hosting rocks.

Kimberlite pipe clusters are confined along the borders of local conductive zones within the kimberlite fields. These conductive zones are connected to the graben type structures, while their borders coincide with the basement and deep sedimentary cover horizon flexure bends, characterized by the heightened fracturing and favorable conditions for kimberlite magma penetration. According to AMT data local conductive zones are surely located and mapped as the most prospective sites for the pipes clusters determination.

The current possibilities of AMT method application for the kimberlite fields and pipe clusters mapping show the necessity of further investigations, focused on the AMT field data obtained over rather large areas for the reliable estimation of these features and their differences from the false anomalies. It should be mentioned that kimberlite fields and pipe clusters have some common regularities as well as their own specific features connected to the geological and geophysical features of the region.

AMT data allow us to outline the faults zones and separate tectonic faults in diamond regions. The kimberlite controlling faults which are usually tilled with trap dykes (localized by the magnetic surveys data) and kimberlite hosting faults in which the trap dykes are absent (which can not be delineated by the magnetic survey) are mapped. AMT data obtained in Nakyn field showed the mapping possibility of cleavage cracks which are connected with the shift deformations in the kimberlite hosting fault zone.

The results suggest that the AMT method is suitable for mapping the kimberlite controlling and kimberlite hosting faults, due to suitable investigation depth (up to a few km) and tensor measurements. However, it is an important to collect data that can create an “image” of kimberlite hosting faults associated with different kimberlite fields.

Location of pipes in difficult geological conditions (the sites covered with traps and thick sediments) at exploration for large and multiphase pipes composed with several kimberlite types including the conductive porphyric kimberlite is demonstrated. The most suitable conditions for the AMT method application are the sites where the traps directly overlie the kimberlite hooting rocks. In this case the mineralogical methods of kimberlites searches are of low value because the overlapping sedimentary rocks and the weathered carbonate ones do not contain indicator minerals. AMT method can be helpful in delineating structural features of pipes and different phases within them. In order to determine the shallow heterogeneity a combination with the low depth electromagnetic survey methods is useful.

Acknowledgements

The authors thank the《ALROSA》JSC staff (V. M. Fomin, M. N. Garat, N. A. Iost, A. V. Manakov, Y. V. Olenius, Y. G. Podmogov, R. F. Salikhov, S. V. Slesarevich, V. M. Zhandalinov, Y. M. Zyuzin) for fruitful collaboration and assistance in organizing and fulfillment of the investigations. We thank B. G. Sapozhnikov and an anonymous reviewer for their useful advices and help in the manuscript preparation.