Introduction

The Deccan flood basalt province of India is one of the largest lava plateau in the world with a present day aerial extent of 500,000 km2 (Mahoney 1988). The volcanics attain a maximum thickness of over 3000 m in the sections of Western Ghats along the Indian west coast (Courtillot and Renne 2003). The origin of the Deccan Volcanic Province (DVP) is still debatable. Many workers believed in the plume origin of this volcanism (Morgan 1981; Richards et al. 1989; Campbell and Griffiths 1990), the alternative suggestion is that this volcanism was related to the continental rift zone (Sheth 2005). Whilst there has been a growing consensus as to the genesis of DVP, the alternative model has remained unclear. Numerous dyke swarms within the DVP occur along the west coast of India (Fig. 1) either parallel to the N-S trending Cambay rift or the E-W trending Narmada-Tapati-Satpura lineament (Fig. 1a), only a few dykes deviate from these trends and oriented in NW–SE and NE-SW directions (Deshmukh and Sehgal 1988). These dykes play an important role in determining the magma emplacement mechanism. Anisotropy of magnetic susceptibility (AMS) is one such study that is used to understand the magma flow pattern of dyke swarms.

Fig. 1
figure 1

a Simplified map of peninsular India showing the major cratons and rift zones (modified after Sheth 2005). b Site map showing 33 dyke locations sampled at west coast south, north and NE of Mumbai along with Panvel flexure and flow stratigraphy of Deccan traps (Salsette, Wai, Lonavala and Kalsubai) of Western Ghats escarpment. The black bar represent studied dykes, red lines faults as per Geological Survey of India (District Resource Maps 2001)

Full size image

AMS is the second-rank tensor referred to as the susceptibility tensor. It is defined by the intensity of the applied field (H) to the acquired magnetization (M) of a material through the equation: Mi = KijHj (Hrouda 1982; Raposo et al. 2004). AMS tensor can be expressed by three principal axes representing maximum (K1), intermediate (K2), and minimum (K3) susceptibility axes. AMS represents the directional variation of magnetic susceptibility within a material and constitutes the contribution of dia-, para- and ferrimagnetic minerals. The magnetic lineation is represented by the K1 axis while the pole of the magnetic foliation (the plane formed by K1 and K2 axes) is K3. AMS in rocks carried by Fe-bearing silicate paramagnetic matrix minerals is due to the magnetocrystalline anisotropy and that of ferrimagnetic minerals, AMS corresponds to the shape anisotropy of these minerals. The study of AMS is the most efficient tool to investigate the problems related to petrofabric orientation in rocks, sedimentology, tectonics and igneous process (Khan 1962; Hrouda 1982; Knight and Walker 1988; Rochette et al. 1992; Raposo and D’Argella-Filho 2000; Chadima et al. 2009; Canon-Tapia and Herrero-Bervera 2009). AMS is a complex phenomenon due to the mixed contributions of magnetic minerals and their domain states to the overall anisotropy of a sample (Rochette et al. 1992). The AMS study of dyke swarms has been found to be a very useful tool in determining magma emplacement kinematics (Knight and Walker 1988; Rapalini and Luchi 2000; Roposo et al. 2004).

Throughout the world, AMS studies contributed much to understanding the flow pattern in dyke swarms (Ernst and Baragar 1992; Raposo and Ernesto 1995; Curtis et al. 2008; Pratheesh et al. 2011; Pan et al. 2014; Kumar et al. 2015; Ramesh et al. 2020; Das et al. 2021). Several studies approached to differentiate sedimentary and tectonic fabric in deformed rocks (Tarling and Hrouda 1993; Parés et al. 1999; Borradaile and Jackson 2004; Levi et al. 2014; Maffione et al. 2015; Weinberger et al. 2017; Bradak et al. 2019) and to characterize earthquake- induced deformation features (Levi et al. 2006; Morner and Sun 2008; Font et al. 2010; Lakshmi et al. 2015, 2020; Chao et al. 2017). In India, AMS studies were applied to determine the magma flow direction in dykes (Prasad et al. 1999; Pratheesh et al. 2011; Kumar et al. 2015; Ramesh et al. 2020; Das et al. 2021) and to another geological context also (Nagaraju et al 2008; Mallik et al. 2009; Mamtani et al. 2013; Renjith et al. 2016).

The dyke swarms in the DVP spread over Maharashtra, Gujarat and Madhya Pradesh. Paleomagnetic and geochemical studies on dykes in India have been carried out extensively but AMS on dykes is scanty (e.g. Vandamme et al. 1991; Radhakrishna and Joseph 1993; Powar and Vadetwar 1995; Prasad et al. 1996; Subbarao et al. 1998; Courtillot et al. 2000; Rao 2002; Srivastava 2006; Ray et al. 2007; Chenet et al. 2008). Patil and Arora (2003) published paleomagnetic results of six dykes from Murud, Mumbai. Basavaiah et al. (2018) revised and reported new paleomagnetic results on 33 dykes from the west coast south, north and NE of Mumbai, Maharashtra. Out of 33 dykes investigated, 29 dykes have yielded stable characteristic remanent magnetizations (ChRM) amenable for statistical analysis. Twenty dykes exhibit N-polarity and nine dykes show R-polarity. This study, however, does indicate the possible presence of two more reversals beyond well-established three-Chron magnetostratigraphy. However, no study on AMS of west coast dykes has been investigated so far. In the present study, the AMS was used for the same 33 dykes from Basavaiah et al. (2018), to investigate the significance of magnetic fabrics. Additionally we also provide information on their mode of emplacement and to understand magma flow direction.

Geological setting and sampling

The dyke swarm outcrop in the DVP namely the ENE-WSW trending Narmada-Satpura-Tapi region containing thousands of dykes, the NNW-SSE trending Konkan or west coast dyke swarms and the Western Ghats swarm NE of Mumbai (e.g., Dessai and Viegas 1995; Bondre et al. 2006). These zones are believed to be regions for mafic dyke swarms (e.g., Sheth 2000). Mafic dyke swarms cover areas of 87,000 km2 in the West Coast belts in the Deccan volcanic province (Deshmukh and Sehgal 1988). The coast-parallel N-S trending dyke swarm extends over 90 km from Mumbai to Murud. These dykes are mainly dolerites of tholeiitic character. The Panvel flexure formed along the west coast as a consequence of late-stage east–west extension that culminated in the post-Deccan rifting and separation of the Seychelles microcontinent (e.g. Dessai and Bertrand 1995; Sheth 1999; Hooper et al. 2010). The area is predominantly occupied by tholeiitic basalts that have been classified as upper Traps (Pascoe 1964). However, from the chemostratigraphic work (Bean et al. 1986) these rocks are included under Poladpur and Ambenali formations of the youngest Wai sub-group of the Deccan basalt group (Subbarao and Hooper 1988) of late Cretaceous to Eocene age (Mahoney 1988). Powar and Vadetwar (1995) suggested that both dykes and flows represent the Poladpur magma type. They confirmed, based on the spatial distribution of the dykes, their close-spaced occurrence, and often restricted thickness that the dykes are hypabyssal intrusives rather than feeders. They also observed that the dykes were emplaced immediately after the outpouring of basalts of Poladpur formation. Based on the field observations, it is suggested that the N-S dykes represent the youngest intrusive phase, while the E-W, NW–SE, and NE-SW dykes are the older intrusive phases within the DVP along the west coast of India.

A total of 33 dykes, west coast south, north, and NE of Mumbai (Fig. 1b, Table 1) were sampled for rock magnetic and AMS studies. The majority of dykes showed tilt angles ranging from 1 to 15°, while few dykes exhibit tilt ∼20° (Table 1). Altogether 158 cores, from 33 dykes mostly from the central part of the dykes were drilled in the field using a portable gasoline-powered drill fitted with a water-cooled diamond drill bit (Stihl, USA). The orientation of the core (i.e. azimuth and dip) is determined with an accuracy of ± 2° using a magnetic compass. The orientation device has a non-magnetic slotted tube with an adjustable platform on which a magnetic compass and inclinometer are fitted. A total of 349 standard cylindrical specimens of size 2.5 cm diameter and 2.2 cm length were cut in the laboratory. In the AMS study, magnetic mineralogy and its orientation is a primary step to understanding the type of fabric and mode of flow. We have collected new and fresh samples for AMS and rock magnetic studies and the results are presented in this study.

Table 1 Site location of west coast dykes, Mumbai
Full size table

Experimental details

The measurement of low-field (200 Am−1 at 976 Hz) AMS for each specimen was carried out using a MFK1 kappabridge with measurements in 64 directions on three mutually orthogonal planes, using an automatic rotator sample holder. The azimuths and magnitudes of principal susceptibility axes (Kmax, Kint, and Kmin) were calculated using SUFAR software supplied by AGICO, together with other magnetic anisotropy parameters such as anisotropy ratios, expressed as corrected degree of anisotropy (P′) and shape (T) (Jelínek 1981). The analysis of the AMS data was performed using the Anisoft 5 software. Detailed rock magnetic experiments were carried out on representative specimens from each dyke in order to determine their main magnetic carrier. Selected samples of each profile were subjected to high-temperature magnetization and hysteresis loop measurements in order to gain further insights into magnetic mineralogy and grain size. Measurements of temperature-dependent susceptibility (κ-T curves) were obtained using AGICO KLY-4S Kappabridge. The samples were heated (from 40 to 700 °C) and cooled back (to room temperature) in an argon atmosphere to reduce the formation of secondary magnetite or hematite. Low temperature (from about −196 °C to room temperature) κ-T curves for two samples from representative dykes were also obtained using a CS3-L apparatus coupled to the KLY-4S bridge instrument. Hysteresis measurements were carried out using a MicroMag Alternating Gradient Magnetometer (AGM) and Molspin Nuvo vibrating sample magnetometer in field of ± 1 T. Values of saturation magnetization (Ms), saturation remanent magnetization (Mrs), coercive force (Hc) and the coercivity of remanence (Hcr) were calculated from the hysteresis loops. All laboratory measurements were carried out at the Indian Institute of Geomagnetism (IIG), Navi Mumbai.

Results

Anisotropy of magnetic susceptibility (AMS)

Anisotropy of magnetic susceptibility (AMS) measurements were carried out on 349 specimens from the samples of 33 dykes selecting not less than four specimens from different samples of each dyke. These measurements were made on the fresh specimens. The AMS data for the dykes is presented in the Table 2. The mean magnetic susceptibility (Km) = (K1 + K2 + K3)/3 in SI units, is overall high and values range between 1.09 × 10–2 and 11.15 × 10–2 SI for present studied dykes (Table 2; Fig. 2a). The degree of anisotropy (P), given by P = K1/K3, from 1.0 to 1.5, as anticipated for basaltic rocks and values range between 1.0 and 1.3 (Fig. 2b) with an average value of 1.10 (Table 2). For the dykes with different fabric types, there is no clear relationship between P and Km parameters (Fig. 2c). Figure 2d shows the P versus T graph (Jelinek 1981), T is expressed by T = (ln F − ln L)/(ln L − ln F) where F = K2/K3 and L = K1/K2. The oblate (T > 0) ellipsoid shape is more predominant in the dykes even though a subordinate group is plotted in the prolate (T < 0) field (Fig. 2d).

Table 2 Anisotropy of magnetic susceptibility data for the studied dykes
Full size table
Fig. 2
figure 2

a Histogram of the mean susceptibility (Km) values; b Histogram of the degree of anisotropy (P) values; c P versus Km and d P versus Jelinek’s parameter (T)

Full size image

The distribution of maximum, intermediate, and minimum susceptibilities at each site-dyke are plotted in Figs. 3–5. The strike of the dyke at each site is indicated for comparison with the anisotropy data. The effect of the dip on the characteristics of the magnetic fabric is insignificant. AMS data from these dykes have been grouped into three (Normal, Inverse and Intermediate) categories. The first category fabric, termed as normal fabric, was characterized by the clustering of K1 and K2 axes on the dyke plane, whereas K3 axes are nearly perpendicular to it (Fig. 3). Second group termed as inverse fabric in which K2 and K3 axes forming a plane parallel to the dyke plane and K1 is perpendicular to that plane (Fig. 4). Third category fabric termed as Intermediate fabric, characterized by K1 and K3 axes clustering close to the dyke plane and the K2 axes are perpendicular to this plane (Fig. 5).

Fig. 3
figure 3

Anisotropy of magnetic susceptibility data of studied dykes plotted in lower hemisphere projections for different fabrics for Normal fabric. Solid squares, triangles and open circles are maximum (K1), Intermediate (K2) and minimum (K3) axes respectively. Dyke trend is shown in yellow line

Full size image
Fig. 4
figure 4

Anisotropy of magnetic susceptibility data of studied dykes plotted in lower hemisphere projections for different fabrics for Inverse fabric. Solid squares, triangles and open circles are maximum (K1), Intermediate (K2) and minimum (K3) axes respectively. Dyke trend is shown in yellow line

Full size image
Fig. 5
figure 5

Anisotropy of magnetic susceptibility data of studied dykes plotted in lower hemisphere projections for different fabrics for Intermediate fabric. Solid squares, triangles and open circles are maximum (K1), Intermediate (K2) and minimum (K3) axes respectively. Dyke trend is shown in yellow line

Full size image

Magnetic mineralogy

Thermomagnetic measurements (κ-T) were obtained for representative samples from dykes within the four AMS fabric types. All κ-T curves in Normal fabric, are characterized by a major decrease in magnetic susceptibility at about curie temperature (Tc) of ~ 550 °C except for KLI1 which shows decrease in susceptibility at Tc 580 °C (Fig. 6a–c). In KLI1, three variations in the slope of the heating curve 480 °C, 540 °C and 580 °C, seems to highlight the presence of either three phases of titanomagnetite or different mineralogical magnetic phases (Fig. 6a). During heating, MRD1 and MRE1 samples show two Tc points, 450 °C for MRD1 and 420 °C for MRE1 (Fig. 6a, b). The susceptibility drops at 450 °C, 420 °C and 580 °C suggests the presence of titanomagnetite and magnetite respectively. In MREI, we observe an increase in susceptibility beyond 450 °C and a rapid decrease towards 550 °C characterizing probably Hopkinson effect or Hopkinson peak. This peak highlights the presence of pure magnetite.

Fig. 6
figure 6

Representative magnetic susceptibility versus temperature (Low and high) curves for samples with different AMS fabrics a-c normal fabric; d-f inverse fabric and g-i intermediate. The red and blue lines are heating and cooling cycles respectively

Full size image

The susceptibility drops of the VHR2 sample, obtained from Inverse fabric, at 330–350 °C and 580 °C indicate the presence of likely Pyrrhotite (Tc⁓320–350 °C) and magnetite respectively (Fig. 6d). Magnetite is probably pure due to the existence of Verwey (1939) transition (around −100 °C) and the Hopkinson peak at 580 °C. The BLR sample shows three Tc at 200 °C, 390 °C and 580 °C indicating likely presence of titanomagnetite and magnetite (Fig. 6e). The KMB sample shows Tc point recorded at 580 °C indicating presence of magnetite (Fig. 6f). KMB has undergone a formation of small quantity of another phase or another magnetic mineral between 450 °C and 500 °C as seen in the slight variations in slope of the cooling and heating curves respectively. KLI2 sample from Intermediate fabric shows two Tc points recorded at 400 °C and 580 °C indicating presence of titanomagnetite and magnetite respectively (Fig. 6g). κ-T curve for the KLI3 sample show Tc at 550 °C corresponds to titanomagnetite (Fig. 6h). The heating curve for KLI3 shows two curie points, the first around 400 °C and the second around 580 °C. The cooling curve has three changes in slope towards approximately 580 °C, 510 °C and 350 °C. This supposes the formation of new magnetic phases. Tc for sample RDA of Intermediate fabric is recorded at 350 °C and 580 °C indicating presence of titanomagnetite and magnetite (Fig. 6i). In this case we have two Tc around 250 °C and 350 °C and both heating and cooling curves are reversible between 700 °C and 550 °C showing that the original amount of magnetite did not undergo mineralogical transformation (Fig. 6d, e). Heating and cooling curves of all specimen are reversible between the highest temperatures and the Curie one. This means that the original magnetite was not altered during heat treatments. The slight transformations that have occurred (reduction or oxidation) have only concerned the other existing phases. These rock magnetic analyzes highlighted the presence of titano-magnetite, magnetite, pyrrhotite and another unidentified mineral. The cooling curves showed the formation of other unidentified magnetic phases or minerals by the transformation of pre-existing minerals.

Hysteresis curves and the parameters on a Day plot (Dunlop 2002) for different fabrics from representative samples from individual dykes are shown in Fig. 7 and Table 3. Values of coercive force (Hc), saturation remanence (Mrs), and saturation magnetization (Ms), obtained at maximum field of 1 T were calculated after subtraction of the paramagnetic contribution. The ratio of saturation remanence to saturation magnetization (Mrs/Ms) and the ratio of remanence coercivity to saturation coercivity (Hcr/Hc) range between 0.02–0.38 and 1.32–6.25 respectively. Hysteresis curves for representative dyke samples that exhibit single domain (SD), pseudo-single domain (PSD) and multi domain (MD) behavior are shown in Fig. 7a–h. Hysteresis parameters data set in the day plot show a classic trend from PSD grains to MD grains, most probably due to a mixture of real PSD and MD grains with similar Ti substitution (Fig. 7i). The representative hysteresis loops are closed mostly around < 100 mT indicating the predominance of the ferromagnetic phases, and all the loops are saturated by 250 mT in an applied field of 1 T (Fig. 7a–h). Thinner loops (KLI1, MRD1, MRE1, BLR, KMB, KLI3 and RDA) are due to low-coercivity components while intermediate (VHR2) suggests the presence of medium coercive magnetic minerals (Fig. 7a–h).

Fig. 7
figure 7

Representative hysteresis loops for the studied dykes a-c normal fabric, d-f inverse fabric g, h intermediate fabric and i hysteresis parameter ratios of Mrs/Ms versus Hcr/Hc for samples from the west coast dykes, Mumbai (after Day et al. 1977) with the boundaries of SD and MD behaviour for magnetite taken from the values of Dunlop (2002). Mrs saturation remanence, Ms saturation magnetization, Hcr remanence coercivity, Hc coercive force. Hysteresis measurement cycles were performed for ± 1 T and in the figure, plotted only for ± 0.5 T for a better view

Full size image
Table 3 Summary of hysteresis measurements of studied dykes
Full size table

We can thus expect for the part of the AMS carried by magnetite, with MD or Pseudo-single domain PSD, a normal magnetic fabric directly related to the shape of the magnetite grains (Potter and Stephenson 1988).

Petrographic studies have been carried out for the same 33 dykes to identify the mineral phases (Basavaiah et al. 2018 for details). Most of the samples show fine-grained basaltic composition containing phenocrysts of subhedral prismatic plagioclase and rare olivine (Fig. 8a–d). The dykes in this area are either dolerite or olivine phyric basalt, or olivine of plagioclase phyric of extremely fine-grained basalt. Dyke MRE2 show Olivine phenocrysts with alteration along margin and interstitial glass within plagioclase laths in groundmass filled with magnetite (Fig. 8a). Dyke KMB shows extremely fine grained basalt with phenocrysts of plagioclase prism and subhedral squarish opaque (Fig. 8b). Dyke MRD2 contains extremely fine grained basalt with elongated crystals of plagioclase as phenocrysts (Fig. 8c). Dyke KLI2 shows phenocrysts of elongated crystals of plagioclase in extremely fine grained basalt (Fig. 8d).

Fig. 8
figure 8

Representative Thin section images for the studied dykes. a MRE2: Olivine phenocrysts with alteration along margin and interstitial glass within plagioclase laths in groundmass filled with magnetite. b KMB: Extremely fine grained basalt with phenocrysts of plagioclase prism and subhedral squarish opaque. c MRD2: Extremely fine grained basalt with elongated crystals of plagioclase as phenocrysts and d KLI2: Phenocrysts of elongated crystals of plagioclase in extremely fine grained basalt. Legend for mineral recognition: Plagioclase (Plg), Olivine (Ol) and Magnetite (Mg). See the text for further explanation

Full size image

Discussions

Normal fabric was observed in 16 dykes and occurred in 48.5% of the dykes (Fig. 3). This kind of fabric was also found in earlier studies of dyke swarms (Rochette et al. 1992; Prasad et al. 1999; Raposo and D’Argella-Filho 2000; Rapaleni and Luchi 2000; Kumar et al. 2015; Ramesh et al. 2020). Normal fabric has been interpreted as a flow fabric with K1 as the flow indicator (Knight and Walker 1988). Several investigators have used the K1 inclination (IK1) of normal fabric to deduce the distance between the fractures and magma source (e.g. Ernst and Baragar 1992; Raposo and Ernesto 1995; Knight and Walker 1988). In dykes with IK1 < 30° is an indication that the dykes were fed by horizontal or sub-horizontal flow (Raposo and D’Argella-Filho 2000). Five dykes appear to be fed by horizontal flow and distributed in the south of Mumbai of Wai Formation. The horizontal magma flow direction revealed by sub-horizontal inclinations in these dykes suggests that the source could be located far away. This type of flow pattern is observed in several dyke swarms (Ernst 1990; Raposo and Ernesto 1995; Hastie et al. 2014; Ramesh et al. 2020). Ray et al. (2008) also assumed about the presence of both inclined to subvertical upward and lateral (although very rare) injection in Central Deccan Traps. Delcamp et al. (2014) reported a similar flow pattern from the mafic dykes of the Tenerife NE rift zone. They compared the upward subvertical flows with the summit eruptions just above the shallow crustal chambers and inclined to distant lateral flow away at the flanks (Njome et al. 2008; Wantim et al. 2011).

The value 30° < IK1 < 60° was assumed to indicate inclined flow and IK1 > 60° indicated the vertical flow. In the present study, seven dykes fed by inclined westward flows and four dykes have the steepest K1 suggests that the region could be closer to a magma source. However, the flow distribution is random and does not show any preferred pattern to suggest the single magma chamber from deep-seated source. In this scenario, the possible interpretation could be the presence of multiple subsurface magma chambers which are responsible for the random distribution. AMS analysis by Das et al. (2021) suggests that the Dhule-Nandurbar Deccan dyke swarm display dominantly subvertical to inclined flow and occasional sub-horizontal/lateral flow. Their study also suggests the presence of multiple sub-surface magma centres from which magma pulses got injected through crustal fissures.

Based on isotopic and geochemical characteristics, Vanderkluysen et al. (2011) and Hooper et al. (2010) inferred that the N–S dykes in the coastal area were a product of post-Deccan Seychelles rifting following the main phase of volcanism, and that the dykes with no preferred orientation in the coastal area were most likely feeders for the three main upper Formations (Fms) of the Wai subgroup (Poladpur, Ambenali and Mahabaleshwar). Moreover, Vanderkluysen et al. (2011) identified the dyke swarm as likely feeders for the lower and middle Fms (Fig. 1b) exhibiting preferred orientations consistent with the rifting based model, whereas the dyke swarms with no preferred orientation inferred to be the feeder dykes of the top Fms are inconsistent with the rifting model. Geophysical model by Bhattacharji et al (2004) reported that the mafic bodies appearing as magma chambers along the western continental margin rift in the upper lithosphere. They are considered as the major reservoirs for the Deccan flood basalt volcanism. Petrological modeling based on olivine clinopyroxene- plagioclase saturated liquid compositions (Grove et al. 1992), using geochemical data on feeder dikes and lowermost Deccan lava flows in the Narmada-Tapti valley and near Surat, also indicates that the Deccan magmas last equilibrated in feeder dikes and associated underlying multiple magma chambers at a depth of about 7 km along the Narmada-Tapti and western continental margin rifts (Bhattacharji et al 1996). 40Ar/39Ar and K–Ar age dating of the feeder dikes and associated lower Deccan lavas indicate that they were coeval and erupted at approximately 65 Ma (Bhattacharji et al 1996). Although no direct physical field evidence of a feeder dyke is found, geophysical, geochemical, and AMS data indirectly proves that the dyke swarm was most likely a feeder dyke swarm to some part of the Deccan flood basalt. Geochemical, petrological and geophysical studies infer the presence of multiple magma chambers at shallow crustal surface (Bhattacharji et al. 1996), which supports flow. The paleomagnetic study carried out by Basavaiah et al. (2018), highlights successive flows at different periods (~ 65 and ~ 80 Ma) with Normal and Reverse polarities. The results show that between these two periods India drifted about 4.4° in altitude. This may indicate that the sources emitting magma are different. In addition, the fact that the dykes have been tilted, the horizontal and vertical distances of the magma emitting sources relative to the outcrop also vary likely.The random distribution of magma in the present study is thus consistent with these conclusions.

The inverse fabric has been observed in ten dykes and occurred in 33% of the dykes (Fig. 4). The inverse fabric in dykes has been interpreted to be due to secondary processes such as post-emplacement modification, hydrothermal alteration, or due to the presence of SD particles in the rocks (Rochette et al. 1992). The hysteresis parameters in the Day plot shows that all the samples fall into the PSD to MD range (Fig. 7i) and are found in other dyke swarms (Tauxe et al. 1998; Raposo and Ernesto 1995). As petrographic analyses showed no evidence either of later alteration due to hydrothermaluids and either metamorphism or solid-state deformation. Alternatively, this Inverse fabric could be related to local irregularities that occurred after dyke emplacement. As seen in Fig. 4, clusters of K2 and K3 in the case of two dykes (KMB and KLI4) are disposed symmetrically on the opposite sides of the dyke trend with an offset of about 30° from the trend. The K1 cluster also is displaced by the same amount from the perpendicularity of the dyke trend. The remaining dykes appear to meet the requirement of this fabric nearly well.

Intermediate fabric, which is characterized by clustering of K1 and K3 axes close to the dyke orientation plane and K2 axes are perpendicular to it (Fig. 5) and is very nearly exhibited by only six dykes. This fabric occurs in 18.2% of the dykes. This kind of fabric was also found in earlier studies of dyke swarms (Rochette et al. 1992; Raposo and D’Argella-Filho 2000; Rapaleni and Luchi 2000). The intermediate fabric has been interpreted to be due to the presence of fine-grained, particularly PSD grains (Rochette et al. 1992; Aubourg et al. 1995). This interpretation cannot be applied to fabric found in present study dykes, since all the three normal, inverse and intermediate fabric samples fall in-to PSD/MD range (Fig. 7i). In the present study, the intermediate fabric in the dykes might be caused due to the vertical compaction of a static magma column with minimum stress along the dyke direction (Park et al. 1988; Raposo and D’Argella-Filho 2000).

Conclusions

The following conclusions can be drawn from the AMS study of 33 dykes from West coast of Maharashtra, DVP:

  1. (1)

    The magnetic mineralogy studies indicate the probable presence of a complex combination of ferrimagnetic grains in the size range PSD/MD. Out of 33 dykes, 27 dykes are dominated by PSD, five are in SD and one in MD.

  2. (2)

    The AMS study has yielded three kinds of magnetic fabric: normal, inverse, and intermediate based on the clustering of K1, K2 and K3 axes with respect to the dyke planes.

  3. (3)

    Normal fabric displays clustering of K1–K2 axes in the dyke plane and K3 axes are normal to the dyke plane. This fabric could reflect the magma flow. Intermediate fabric found in six dykes and was characterized by K1–K3 axes clusters close to dyke plane whereas K2 axes are perpendicular to the dyke plane. Inverse fabric defined by K2–K3 plane parallel to the dyke plane and K1 perpendicular to dyke plane, found in 11 dykes.

  4. (4)

    The inclination of IK1 axes, which gives magma flow direction in dykes displaying normal fabric, dykes were mainly fed by inclined westward plunging flows (30° < IK1 < 60°) to the steepest IK1 (> 60°) suggests that the dykes may be closer to magma source. Horizontal magma flow inferred from three dykes reveals source is located further away.

  5. (5)

    Presence of multiple trends of primary flow axes revealed from AMS study support subsurface magma chambers which are responsible for the random distribution. Subvertical upward flow indicates the proximity of source chamber. The observed flow from the present study together with geophysical, geochemical and petrological evidences provided by previous studies support indirect evidences of the theory of fissure fed volcanism.