You are on page 1of 17

Bull Volcanol (2004) 66:2945 DOI 10.

1007/s00445-003-0294-x

RESEARCH ARTICLE

Ninad R. Bondre Raymond A. Duraiswami Gauri Dole

Morphology and emplacement of flows from the Deccan Volcanic Province, India
Received: 14 December 2001 / Accepted: 20 April 2003 / Published online: 5 July 2003  Springer-Verlag 2003

Abstract The present study is probably the first of its kind in the Deccan Volcanic Province (DVP) that deals in detail with the morphology and emplacement of the Deccan Trap flows, and employs modern terminology and concepts of flow emplacement. We describe in detail the two major types of flows that occur in this province. Compound pahoehoe flows, similar to those in Hawaii and the Columbia River Basalts (CRB) constitute the older stratigraphic Formations. These are thick flows, displaying the entire range of pahoehoe morphology including inflated sheets, hummocky flows, and tumuli. In general, they show the same three-part structure associated with pahoehoe flows from other provinces. However, in contrast to the CRB, pahoehoe lobes in the DVP are smaller, and hummocky flows are quite common. Simple flows occur in the younger Formations and form extensive sheets capped by highly vesicular, weathered crusts, or flow-top breccias. These flows have few analogues in other provinces. Although considered to be aa flows by previous workers, the present study clearly reveals that the simple flows differ considerably from typical aa flows, especially those of the proximal variety. This is very significant in the context of models of flood basalt emplacement. At the same time, they do not display direct evidence of endogenous growth. Understanding the emplacement of these flows will go a long way in determining whether all extensive flows are indeed inflated flows, as has recently been postulated.
Editorial responsibility: T. Druitt N. R. Bondre ()) Department of Geology, Miami University, 114 Shideler Hall, Oxford, Ohio, 45056, USA e-mail: bondren1@muohio.edu Tel.: 513-529-3227 Fax: 513-529-1542 R. A. Duraiswami G. Dole Department of Geology, University of Pune, 411007 Pune, India

Most of the studies relating to the emplacement of Continental Flood Basalt (CFB) lavas have relied on observations of flows from the CRB. Much of the current controversy surrounding the emplacement of CFB flows centers around the comparison of Hawaiian lava flows to those from the CRB. We demonstrate that the DVP displays a variety of lava features that are similar to those from the CRB as well as those from Hawaii. This suggests that there may have been more than one mechanism or style for the emplacement of CFB flows. These need to be taken into account before arriving at any general model for flood basalt emplacement. Keywords DVP Flows Pahoehoe Compound Simple Inflation Emplacement

Introduction
Some recent models for the emplacement of continental flood basalts (CFB) hypothesize that many of these flows are inflated pahoehoe flows, and that they may have been emplaced in a gentle fashion analogous to inflated pahoehoe flows from Hawaii and Iceland (e.g., Thordarson and Self 1998). This has, however, been contended by other studies that propose that CFB flows have been emplaced rapidly (e.g., Anderson et al. 1999). Different models have been proposed to explain the exact nature of endogenous growth in basaltic lava flows and lobes, and aspects such as the time scales and pulses of inflation have been debated (Anderson et al. 2000; Self et al. 2000). These studies have focused only on the Columbia River Basalts (CRB) and active volcanoes such as in Hawaii. In this paper, we seek to integrate the observations and inferences of our ongoing studies in the Deccan Volcanic Province (DVP; Fig. 1). While we have made observations on the northern and eastern parts of the province, we focused on detailed documentation of features from the western part (Fig. 2) where exposures are excellent and most of the previous stratigraphic work has been done.

30 Fig. 1 Map showing the extent of the Deccan Volcanic Province in India (modified after Deshmukh 1988). The distribution of compound pahoehoe and simple flows is also shown

Fig. 2 Map depicting locations where observations were made during the past 3 years. Boxes show regions for which detailed maps are shown in Fig. 3. Several of these have been referred to in the text. The principal rivers and the position of the Western Ghats Escarpment are also shown

31 Fig. 3 Map of geology and localities studied in a the Pune region and b the Sangamner region

We describe lava flow morphology from the DVP by considering examples of each from specific locations within this region. We then discuss the emplacement of flows and formation of associated features, and the implications that they may have for the evolution of the DVP in particular, and CFB provinces in general. Four areas have been the focus of most of our studies. Details of tumuli and small pahoehoe lobes around Dhule and Daund (Fig. 2) have been described elsewhere (Duraiswami et al. 2001, 2002). In this paper, we focus on studies of lava flow morphology from the Pune and Sangamner areas (Fig. 3) in addition to observations from numerous other localities. This work is largely qualitative as a consequence of the lack of even the most basic data on various aspects of the Deccan flows that would have aided a more quantitative approach. In spite of this, we feel that it is a good basis for more comprehensive studies in the near future.

The Deccan Volcanic Province


The DVP is one of the largest CFB provinces in the world, covering an area of more than 500,000 km2 in the western and central parts of India (Fig. 1). It is postulated that an equivalent area of the Deccan Traps has been down-faulted along the western coastal region and the volume of erupted material could have been well over a million km3. At places along the western part of the province, a continuous vertical succession of basaltic

flows with a thickness of more than 1,200 m can be observed, and geophysical studies have indicated that the thickness attained by the lava pile is over 2,000 m in the western part of the province (Kaila 1988). On the fringes of this province around Nagpur (Fig. 1), the lava pile decreases considerably in thickness to as little as 10 m. It has been postulated that the DVP formed around 65 Ma in response to the passage of the Indian plate over the Reunion hotspot (e.g., Beane et al. 1986; Courtillot et al. 1986; Cox and Hawkesworth 1985; Morgan 1972). The eruption of the Deccan lavas is intimately related to the formation of the passive margin along the western coast of India, as the Seychelles micro-plate separated from the Indian plate. The DVP is constituted dominantly of tholeiitic basaltic lava flows, nearly horizontal, stacked one above the other. The sequence exhibits a low dip of about 1 to the southeast. The basaltic flows are intruded at a number of locations by essentially doleritic dykes, some of which are speculated to represent feeders. Occasional alkaline intrusions have also been noted from this province (see Auden 1949; Powar 1987). Considerable work has been done on the Deccan basalts in the past two decades, especially with respect to their stratigraphy, geochemistry, geochronology, and structural geology (e.g., Subbarao 1988, 1999 and references therein). One of the most important achievements of this work has been the establishment of a chemostratigraphy for a large portion of the DVP (e.g., Beane et al. 1986; Cox and Hawkeshworth 1985; Devey and Lightfoot 1986; Mitchell and Widdowson 1991; Subbarao et al.

32 Table 1 Chemostratigraphic classification of the Deccan Trap basalts (Subbarao and Hooper 1988). This classification evolved from concerted efforts on the part of several groups of researchers that mapped the DVP during the early 1980s and 1990s. Each formation refers to a mappable package of lava flows that display similar physical, textural, and geochemical characters, and which may be separated by giant phenocryst basalts. Distinctive flows within each formation have been classified as members (not shown in this table). All formations have been grouped into three subgroups based on some common characters, while these three subgroups together define the Deccan Trap Group. See text for more details Group D E C C A N B A S A L T Subgroup Wai Formation Desur Panhala Mahabaleshwar Ambenali Poladpur Bushe Khandala Bhimashankar Thakurwadi Neral Igatpuri Jawhar

Lonavala Kalsubai

1988). Flows have been grouped into Formations and Subgroups based on their field relations and geochemistrymajor elements, trace elements, and trace element ratios. This stratigraphy (Table 1) is best constrained along the western part of the province, while mapping and geochemical studies are ongoing in the northern and eastern regions (e.g., Malwa Traps and Mandla Lobe; Fig. 1). These studies have indicated that the mapped Formations extend over distances of hundreds of kilometers and occupy areas of the order of tens of thousands of square kilometers. A lithostratigraphic classification of these lavas also exists, as formulated by the Geological Survey of India (e.g., Godbole et al. 1996), which is based entirely on the field characters of the lava flows. Crucial to both these classification schemes are regionally extensive flows containing giant phenocrysts of plagioclasetermed Giant Phenocryst Basalts in the Deccan which have been used to separate the Formations. It was widely believed that the Deccan Traps represent a fissure type of eruption and that sheets of lava erupted through numerous fissures/cracks in the crust and spread laterally over considerable distances, sometimes across hundreds of kilometers (West 1959). Based on the plume model, it has been suggested that the main phase of eruption originated mainly from one major eruption center, which was located in the northwestern part of the province (Beane et al. 1986). The observed southward migration and overstepping of the chemostratigraphic Formations is believed to be a consequence of the northward movement of the Indian Plate over the Reunion hotspot (Watts and Cox 1989; Mitchell and Widdowson 1991). However, several other views regarding the source of the Deccan lavas exist, and multiple eruption centers

have also been proposed (Bhattacharjee et al. 1996; Kale et al. 1992). Some notable work on the Deccan lava flows and their morphology has been done in the past few decades (e.g., Karmarkar 1978; Marathe et al. 1981; Walker 1969; Rajarao et al. 1978; Deshmukh 1988). Walker (1969) made some very important observations on the Deccan lavas and his analysis of various volcanogenic features is one of the best in the province so far. He was the first to introduce the terms Simple and Compound to flows from the Deccan, and these terms are still prevalent, although they have not always been used in their original sense. Karmarkar (1978) and Marathe et al. (1981) discussed the characters and emplacement of flows from the western DVP, describing the flows have been described as Compact, Amygdaloidal, etc, which are terms without any genetic significance. Rajarao et al (1978) provided an informative description of the Deccan lavas and described them as pahoehoe and aa. They also mentioned the occurrence of flows with mixed characters and the fact that aa flows from the Deccan lack the basal breccia horizon. Deshmukh (1988) discussed the general characters of compound pahoehoe flows from the DVP and the mechanisms responsible for petrographic variations in pahoehoe flow lobes. In the past few years, sinuous lava tubes and channels have been reported from some parts of the province (e.g., Thorat et al. 1996; Misra 2002). Dole et al. (2002) have, however, commented on the validity of these observations. In general, though, studies of the physical characters of the Deccan lava flows and their emplacement have developed in isolation from similar studies in other parts of the world. This has resulted in the development of local terminology and has resulted in considerable confusion. Although several workers made detailed observations of lava flow morphology, a plethora of terms have been used and this renders any comparison and meaningful interpretation difficult. Virtually no work has been done on such aspects of the Deccan lava flows and their emplacement as vesicle distribution, segregation structures, viscosity, and effusion rates. Hence, while voluminous information is published regularly on the physical volcanology of lava flows world-wide, this is not the case for the Deccan lavas. Recently, some studies (Bondre et al. 2000; Duraiswami et al. 2001, 2002, 2003) have tried to document the character and emplacement style of flows from the DVP and compare them with those from other provinces. These studies suggest that the pahoehoe flows and certain associated features from the DVP are very similar in nature and scale to those from Hawaii.

Flows from the Deccan Volcanic Province


Although detailed maps of individual flows are not available for most Deccan Trap flows, it is well known that they extend over great distances. For example, Rajarao et al. (1978) mention certain flows that have been seen to extend over areas of 3,000 km2 without pinching

33

Fig. 4 Map showing the extent of two of the chemostratigraphically defined formations in the Deccan Volcanic Province. Other formations have been omitted for clarity

out. As mentioned earlier, the chemostratigraphic formations extend over distances of hundreds of kilometers (Fig. 4). It is possible to treat many of these Formations as individual flow fields, as defined by Keszthelyi et al. (1999). Flows from the DVP can be grouped into two main categories, Compound pahoehoe and Simple flows. The former type is almost exclusively confined to older Formations (Kalsubai Subgroup; Table 1), while the latter dominates in the younger Formations (Wai Subgroup; Table 1). In addition to this vertical disposition, the two types occupy distinct geographical areas (Fig. 1). The area depicted as dominantly compound exposes the older Formations, while that labeled dominantly simple exposes the younger Formations. Walker (1969) suggests that flows are compound close to the source and become simple further away from it. This may be why compound pahoehoe flows in the DVP occur around the proposed source region in the northwest, while the simple flows occupy peripheral positions. However, it has yet to be demonstrated that the simple flows did indeed originate in the northwestern part of the province and flowed to the south and east. Compound pahoehoe flows also occur in the Saurashtra region (Fig. 1), and such flows have been reported from the Mandla Lobe (Fig. 1; Solanki et al. 1996). Compound pahoehoe flows As Prof. G.P.L. Walker observed during his visit to the Deccan (Walker 1969), compound pahoehoe flows have probably been developed on a unique scale in the DVP and this makes it an excellent place to study such flows.

Fig. 5 Logs constructed from observations in wells from the Igatpuri region (Fig. 2). Note the thickness of flow lobes (as defined by a core and crust pair) and the relative thickness of cores and upper crusts of lobes

In spite of the ancient nature of this province, typical pahoehoe features are extremely well-preserved. Exhumation by drainage has acted in such a fashion that at several places it is possible to study the surface morphology of flow lobes. Delicate ropy structures and festoons observed in the DVP are identical to similar features in very young volcanoes. Each Formation of the Kalsubai Subgroup, and the Bushe Formation of the Lonavla Subgroup, consist of a sequence of thick (>150 m at places) compound pahoehoe flows. Compound pahoehoe flows are constituted of multiple flow lobes (flow units; Nichols 1936) of varying dimensions. In the DVP, although individual flows are often >100 m thick (Walker 1969; Deshmukh 1988), they are strongly compounded on a local scale, similar to their Hawaiian counterparts (Bondre et al. 2000; Duraiswami et al. 2001, 2002). Walker (1969) measured 31 flow lobes in 120 m thickness of a compound pahoehoe flow in the Igatpuri region, giving an average thickness of 3.7 m per lobe (ranging from 0.3 to13 m). Observations made in wells from the Igatpuri region corroborate these observations (Fig. 5). Observations of these flows in the western DVP can be made both in vertical sections and on surface exposures (exhumed). Studies of vertical sections of flow lobes are facilitated by stream and road cuts, while excellent exposures of exhumed surfaces are observed along the channels and banks of streams. Some of the constraints on observations include discontinuous out-

34

Fig. 6. a Field sketch of a vertical section of a compound pahoehoe flow observed at Warje (Fig. 3a). Note the typical subdivision of flow lobes into upper crust, core, and lower crust. The figure also shows part of a simple flow that rests on the underlying compound pahoehoe flow. b Field sketch of a vertical section through P-type flow lobes observed east of Ganore (Fig. 3b). c Sketch of flow lobes exposed along the walls of the gorge of the Mahalungi River at

Nimgaon Bhojapur (Fig. 3b). Note the termination of lobe 2 against lobe 4. Only the upper crust of lobe 4 is exposed, while the basal crust, crudely jointed and spheroidally weathered core, and part of upper crust of lobe 1 are exposed. d Sketch of small, bun-shaped pahoehoe lobes exposed along the walls of a gorge at Dapur (Fig. 3b)

crops due to highly dissected terrain, difficulty in observing and accessing flows exposed along high vertical cliffs, and thick soil mantle and lack of exposure in some regions. We have documented characters of these flows from the Dhule, Daund, Sangamner, and Pune areas of the DVP. As mentioned before, we focus in this paper on the latter two areas. In regions of relatively higher reliefe.g., around Sangamnerthe thicker flows form vertical to subvertical cliff faces. Natural caves often form by weathering along the boundaries between flow lobes (Bondre et al. 2000). In areas of relatively low relief, such as around Pune city, these flows can be studied in a number of isolated hills. Each flow in vertical section characteristically consists of a few thick (! 10 m), laterally extensive (up to 100 m observed in field exposures) sheet lobes, and numerous much smaller lobes of restricted lateral extent. The larger lobes (and many smaller ones too) display a characteristic three-tiered structure (Aubele et al. 1988; Thordarson and

Self 1998) with a thick, vesicular upper crust, a dense core, and a thin lower crust (Fig. 6a). The tops of many of the smaller lobes display oxidized glassy rinds and ropes (Fig. 7a), while pipe amygdales (pipe vesicles filled with secondary silica) are almost always present at the base (Fig. 7b). Lobes can be classified as either S-type or as Ptype lobes of Wilmoth and Walker (1993). S-type lobes lack pipe vesicles and are vesicular throughout their vertical extent. P-type lobes are characterized by pipe vesicles and display a typical internal structure with a vesicular base and top, and a relatively vesicle poor core (Fig. 6a, b, c). In the DVP, flow lobes of S-type are less common than the P-type lobes. Table 2 depicts representative measurements of sheet lobes from the western DVP, while Table 3 shows the dimensions of smaller lobes measured in a road cut in Karhe Ghat (Fig. 3). Typical hummocky flows with tumuli are also observed in several parts of the DVP (Fig. 7c), especially along the channels and banks of rivers where they have been

35

Fig. 7. a Ropy structure observed at Kondhwa Hill (Fig. 3a). b Pipe vesicles exposed east of Ganore (Fig. 3b). c Hummocky pahoehoe flow exposed at Daund (Fig. 2). d Squeeze-up occupying the inflation cleft of a lobe at Kondhwa Hill (Fig. 3a)

Table 2 Data on representative sheet lobes from different parts of the western Deccan Volcanic Province Location Pune University hill Total thickness (m) 19 Crust (m) 6 Core (m) 13 Remarks Base not exposed. Median horizontal joint plane well exposed. Core with segregation veins, rare vesicle cylinders. Extensive in length (>200 m). Fairly extensive sheet lobes with segregation features (>50 m in length exposed in the section)

Warje

5.5 7.0 4 2.05 12

3.1 2.0 1.70 1.20 6

2.4 5.0 2.30 0.85 6

Kondhwa hill

Thick, extensive (exposed length >50 m) sheet lobe. Crust with numerous squeeze-ups and some tumuli. Squeeze-ups intruded horizontally into the crust and have twisted forms. Pipe-like vesicle cylinders. Base is smooth and horizontal, contact with lower flow marked by red horizon Some brecciated patches are observed in the otherwise smooth crust of this sheet lobe. Segregation veins display branching patterns. Gradational between sheet lobe and tumulus. Fairly steep dips of crustal slabs observed Segregation veins are found at four different levels and show branching. Numerous vesicle cylinders are present; domed vesicles are found at 2 different levels Lower surface is hummocky in nature

Katraj Ghat South of Sangamner

6.9 2.12

5 0.92

1.9 1.0

Northwest of Sangamner South of Sangamner Karhe Ghat

7.5 7

4.50 5

3.0 2 3.1

Partially exposed 6.5

South of Sangamner

Partially exposed 2.5

5.5

36 Table 3 Measurements of dimensions of smaller lobes observed in a road cut in Karhe Ghat. The section is about 10 m thick and consists of a couple of sheet lobes and numerous overlapping smaller lobes.P and S refer to P-type and S-type lobes, respectively. The preponderance of relatively small P-type lobes is evident. Length of lobe (m) 4.9 0.95 0.2 3.9 2.2 Partially exposed 8.6 22.5 0.54 4.45 6.78 1.9 4.8 1.7 1.65 3.75 0.2 3.2 0.3 2.6 6.2 3 1.5 Thickness of lobe (m) 0.8 0.28 0.12 0.54 0.44 2.68 1.02 1.73 0.3 1 0.8 0.3 0.95 0.95 0.6 1.1 0.12 2.15 0.2 0.5 1 0.5 1.5 Description No distinct core, vesicles present throughout Pipes are present at the distal end of the toe No pipes present, unit is entirely glassy Type S P-type S-type P P P P P P-type S S S P P P P P-type P P-type P P P P

toe toe

Pipes are present at the distal end of the toe Highly undulating lower surface

toe

toe toe

Associated 10 p-type break-outs Associated 1 s-type and 9 p-type break-outs

exposed due to exhumation. The tumuli in the DVP are very similar to the Hawaiian tumuli in morphology and dimensions (Duraiswami et al. 2001, 2002). Tumuli in vertical sections can be recognized by the virtue of steeply dipping crustal slabs and inflation clefts occupied by squeeze-ups (Fig. 7d) or crack fillings. Although no systematic studies documenting textural and petrographic variations across pahoehoe lobes have been undertaken, the petrography of a few lobes from the present study reveal interesting insights into textural variations and cooling history of the lava. The outermost 1 to 3 cm quenched selvage of the toes and the lobes show a predominantly hypohaline texture. A near total absence of microphenocrysts and the absence of opaques (ilmenite/titanomagnetite) characterize this zone (also see Thordarson and Self 1998). The upper crust and the basal crust exhibit a typical hypocrystalline texture with fine FeTi oxide minerals. The core tends to be more crystalline but small intra-mineral voids lend a diktytaxitic texture to the core. Intergranular, ophitic and subophitic textures are often seen in thick sheet-lobes. In the core, FeTi oxides are seen as subhedral skeletal crystals. Interesting petrographic variations in tumuli have been discussed by Duraiswami et al. (2001). We measured vesicle sizes and documented their distribution in more than 20 lobes in the western DVP. The cores of Deccan lobes are very sparsely vesicular and hence only vesicles from the upper and lower crusts provide any meaningful data. In most cases the vesicle parameters were measured in vertical sections or in hand samples, while in some cases, exhumed surfaces were used. Measurements were made at numerous levels within individual flow lobes. Wherever distinct vesicle banding was observed, vesicles in those bands were measured. At each level, the long and short axes (L1 and L2) of around

50 vesicles were measured using a vernier calliper. Vesicle frequency (F) was determined by counting all vesicles within areas marked on the outcrops or on hand specimens, and calculating the means of these. It is likely that in some cases, the true three-dimensional nature of the vesicles was not documented. In the future, we hope to improve upon this by measuring vesicles by using sections oriented at different angles within the same lobe, and other more advanced techniques. Implications of these patterns are discussed in the next section, while a more detailed account of vesicle distribution studies from the western DVP is currently under preparation. The characters of each of the three divisions, namely the upper crust, core and the lower crust of lobes, sheets, and tumuli are described in detail below. Upper crust The thickness of the upper crust ranges from about a third to half the thickness of the lobe, and it commonly displays vesicle layering. The largest sheet lobes that we have measured have crustal thickness of around 10 m. Fig. 8 shows the observed vesicle distribution from the crusts of two lobes exposed in the Fergusson College Hill, Pune (Fig. 2a). Lobe 6 displays a downward increase in the average vesicle diameter [(L1+L2)/2], and a dramatic decrease in the frequency (F; number of vesicles/cm2), similar to observations by Cashman and Kauahikaua (1997). Lobe 4 shows a similar trend, although the pattern for F versus depth below the top suggests alternating vesicle-rich and vesicle-poor bands. This is corroborated by field observations of the crust of the flow lobe. The elongation ratio (E=L1/ L2; ratio of long axis of vesicle to the short axis) appears to vary according to the thickness

37 Fig. 8 Vesicle distribution patterns from two lobes exposed at the Fergusson College Hill. Average vesicle size [(L1+L2)/ 2], average frequency (F) and elongation ratio (E) have been plotted against D, the distance in centimeters below the exposed upper surface of the flow lobe. L1 and L2 refer to the long and short axes of a vesicle, average frequency is the number of vesicles per unit area, while the elongation ratio equals L1/L2

of the lobe. In thicker lobes such as Lobe 4 and Lobe 6, E increases slightly with depth below the top, and may increase dramatically near the transition with the core. The transition from the crust to the core in such cases is characterized by some large (>2 cm) flattened, elongated or domed vesicles. In thinner lobes (~ 3 m), E decreases markedly with distance below the top. In another paper (Duraiswami et al. 2003) we have described the vesicle distribution and deformation in a slabby pahoehoe flow, for which vesicle deformation due to shear was much more significant. Crusts of pahoehoe sheet lobes from the CRB typically display considerable vertical, hackly jointing (e.g., Thordarson and Self 1998). In contrast to this, well-developed vertical jointing is rarely observed in crusts of the Deccan lobes. If present, vertical jointing is poorly developed. A sheeted appearance due to differential weathering of vesicle-rich and vesicle-poor bands is more common and in such cases, this zone is extensively calcretized. Anderson et al. (1999) describe the jointing of Hawaiian tumuli and inflated sheet lobes as consisting of three distinct zones: an upper zone of crude to well-defined columnar joints, a middle zone of planar fracture surfaces with no distinct fracture surface features, and a lower zone of fracture surfaces that show evidences of brittle and ductile deformation. Some tumuli in the Deccan (e.g., Daund) were seen to exhibit a similar structure, but most lobe crusts are poorly jointed. Thick squeeze-ups often break through the crust and sometimes form small pahoehoe toes. They may occupy axial clefts of tumuli on the crusts of the lobes.

Sometimes, offshoots of squeeze-ups are seen to have intruded horizontally into the crust (e.g., Fig. 5, Duraiswami et al. 2001). Wherever exhumed surfaces are seen, as along river channels, they are usually gently hummocky in nature and riddled with squeeze-ups that may be as much as a meter wide. We are currently involved in generating high-scale surface maps as well as documenting the morphology of these features in vertical sections. The crust has a much finer texture as compared to that of the core, with glass forming a significant component of the groundmass. The crusts of smaller lobes and toes are essentially thin glassy rinds, often displaying a red to orange color due to oxidation and alteration. The crusts of some lobes are characterized by the presence of patches of breccia, or slabs. Such lobes represent transitional pahoehoe, but are relatively minor constituents of the Deccan flows. In another paper (Duraiswami et al. 2003), we describe a slabby pahoehoe flow transitional to aa and discuss its emplacement and implications. Core The interface between the crust and the core in thick lobes is usually sharp, marked by a change in the style of jointing and the distribution of vesicles. The core has a dense appearance and is sparsely vesicular. It rarely displays well-developed columnar jointing, while crude, four-sided columns or irregular joints are more common. Spheroidal weathering is commonly observed along these

38

joints. Different types of segregation structures are observed in the core. Flow lobes around Pune and in the area around Sangamner afford detailed observations to be made on these features. The segregations are usually darker than the host and slightly coarser in texture. Vesicle cylinders (Goff, 1996) are seen in the lower and middle parts of the core, while segregation veins (vesicle sheets; Thordarson and Self 1998) occur near the interface between the crust and the core. Figure 9a is a sketch of part of a flow lobe containing segregation veins from the Karhe Ghat near Sangamner. Most veins are up to 5 cm in thickness, but thicker (>25 cm) veins are not uncommon. The segregation veins display an irregular form and branching patterns, which indicates that they occupy joints formed at the early stages of cooling. It is also possible that they propagated by hydraulic fracturing (Walker 1993). Their sinuous nature may also indicate that the viscous lava during later stages of cooling was still mobile, which led to the deformation of the veins. The vesicle cylinders are almost always rootless, and are not connected with the pipe vesicles in the basal crust, unlike in the CRB (Thordarson and Self 1998). A typical cylinder in the Deccan is broader in the middle and tapers towards the top and the bottom. However, mushroomshaped and pipe-like forms have also been observed (around Kurundwad and at Kondhwa, respectively). Cylinders in thick lobes can be up to 1 m tall. They are sometimes observed to connect to segregation veins. In the cores of some flow lobes, these structures are essentially trails of vesicles and are not occupied by segregated material. Cylinders are also often bent or at an angle to the flow-top, probably in the direction of flow. The core usually displays ophitic or subophitic textures. Basal crust The basal crust generally does not vary greatly in thickness irrespective of the thickness of the lobe (except in very small lobes). A measurement of the thickness of the basal crust from 46 lobes indicates that the mean thickness is around 0.3 m. This may, in part, have been a result of a sampling bias in favor of smaller lobes that have thinner basal crusts or have only a glassy rind. In the thicker lobes (>3 m), the thickness of this zone is usually around 0.5 m. The lowermost portion of this zone is chilled and glassy. Pipe vesicles (almost always showing an inverted Y geometry) filled by cryptocrystalline silica are very common (Fig. 7b). Pipe vesicles/amygdales are observed even in the smallest lobes. Although many workers consider pipe vesicles as belonging to the same category as vesicle cylinders, those in the Deccan never contain segregated material. Small (around 2 mm), more or less spherical vesicles also occur in this zone. Measurements on two lobes (Lobe 1 and Lobe 2) from the Fergusson College Hill (Fig. 3a) yield values of 2.28 and 3.24 mm for average size, 1.4 and 1.37 for E, and 2/ cm2 and 3/cm2 for F, respectively. Pipe vesicles are not always restricted to the base but sometimes extend well

Fig. 9. a Sketch of part of a flow lobe at Karhe Ghat (Fig. 3b). A, B, C, and D refer to the crust, zone of elongated vesicles, jointed core, and segregation veins respectively. b Sketch of a thick sheet lobe observed at Kondhwa Hill (Fig. 3a). Note the relatively sharp and flat contact of the flow lobe with underlying lobes. c Sketch of a lobe exposed to the southwest of Sangamner (Fig. 3b). The undulating basal contact is clearly visible

into the lower core. Pipe vesicles occurring at two different levels are not uncommon and excellent examples are seen in the Karhe Ghat near Sangamner (see also Fig. 7b). These features are indicative of multiple injections of lava and suggest endogenous growth (Wilmoth and Walker 1993). Recently, we have observed spiracles at the base of a flow lobe northwest of Pune that have incorporated clay, and this is being documented in detail. The geometry of the contact with underlying lobes depends primarily on the microtopography on which the lobe/lobes were emplaced, as described by Duraiswami et

39

al. (2002) for flow lobe tumuli. This is underlined by the fact that in two lobes that are vertically superposed, the pipe vesicles may bend in completely different directions. Larger sheet lobes are usually characterized by planar bases (Fig. 9b); however, the contact can be quite undulating if they are underlain by hummocky flows (Fig. 9c). Simple flows The term simple flow was first coined by Prof. G.P.L. Walker during his visit to the Deccan in 1969 (Walker 1969, 1971), and has since been used to describe the second major category of flows in the DVP. As defined by Walker (1971), the term simple refers to flows that are themselves not constituted of smaller flow lobes. These flows occur in the northern, southern and eastern parts of the DVP where compound pahoehoe flows are rarely observed (Fig. 1). This includes the younger stratigraphic Formations (Khandala Fm of the Lonavla Subgroup and all Formations of the Wai Subgroup). Simple flows vary in thickness generally from a few meters to several tens of meters. Flows greater than a hundred meters thick have been reported from the northern parts of the province (Rajarao et al. 1978). Thick exposed sections of simple flows are observed in regions such as Mahabaleshwar and Toranmal. Mahoney et al. (2000) describe a 870-m-thick section from the Toranmal area, consisting of simple flows ranging in thickness from 10 to 120 m. Most such studies have, however, focused on the chemostratigraphy of the flows, and detailed descriptions of simple flows are sparse. Each simple flow tends to maintain a relatively constant thickness over considerable distances (Fig. 10a, b). It is difficult to quantify these distances in view of the dissection in the province as well as poor exposure in most of the eastern parts. An example of the distances involved is a simple flow that can be observed for most of the way from Pune to Daund (approximately 80 km). Each simple flow appears to be a single cooling unit and there is presently no evidence to indicate the presence of multiple flow lobes. Keszthelyi et al. (1999) state that most simple flows may actually be large sheet lobes which, when traced over a considerable distance, will terminate against other lobes. Although this observation has been verified for some of the CRB flows, we have yet to observe something of this nature in the DVP. In the absence of a suitable alternative term to describe these flows (given that they are sufficiently different from pahoehoe flows from the Deccan), they have been referred to as simple flows in this paper. We made detailed observations of several outcrops in the Pune areaKatraj Ghat and Dive Ghat; flows belonging to the Poladpur and Ambenali Formationsand around Satara and Kurundwadflows belonging to the Poladpur, Ambenali, and Mahabaleshwar Formations. The thickness of flows ranges from 10 to 50 m, and flows can be followed along cliffs in the vicinity for kilometers. Figure 11

Fig. 10. a Photo showing a laterally continuous simple flow (dark band towards the top marked by arrow) observed close to Satara (Fig. 2). b Simple flows seen along cliffs in the Satara region (Fig. 2). The dark bands are thickly vegetated flows

displays the internal structure of a typical simple flow from the DVP, based on our observations from the western DVP and observations of previous workers from this and other regions. A typical simple flow displays three zones crustal, central, and basal as described below. Crustal zone Our observations suggest that the nature of the crustal zone of simple flows varies considerably. Although some flows do display preserved crusts, many are characterized by reddened (oxidized), rubbly tops (Fig. 12a) or flow-top breccia, as observed around Kurundwad and Satara. The highly vesicular/amygdaloidal nature (almost scoriaceous at places) and fine to glassy texture are immediately apparent. The frequency of vesicles appears to be high throughout the crust, unlike the crusts of inflated pahoehoe lobes where it decreases with depth below the top. The vesicles are also much smaller (~12 mm). Quantitative studies on vesicle distribution patterns of these flows are ongoing. Lava blisters, filled with cryptocrystalline silica are sometimes observed. Small lobes and

40

Fig. 11 Morphology of a typical simple flow. The upper parts of the upper flow, as well as the base of the lower flow have not been depicted, since this is how such flows are exposed in most lowrelief regions

squeeze-ups that are ubiquitous in the pahoehoe flows, are almost never observed in the crustal zone of the simple flows. The flow-top breccia is usually constituted of irregular, highly vesicular/amygdaloidal fragments in a matrix of fine, often glassy material (Fig. 12a) and ranges from 0.85 to 5.00 m thick. This zone is invariably reddened and often takes on a crumbly appearance. The flow-top breccia grades into the underlying flow and is rarely entrained within the coherent part of the flow. The general appearance is of a well-developed, vesicular crust that has been disrupted at some stage in the emplacement of the flow. It is very difficult to trace a single flow and to ascertain if the flow-top breccia is a continuous horizon, or if it grades into something else. Along several cliff sections (e.g., Mahabaleshwar), however, the contact between two flows is clearly visible for a considerable distance due to the differential weathering of flow tops. This topographic expression probably owes its origin to the softer flow-top breccia and suggests that such horizons are fairly continuous. Central zone This is usually the thickest part of the flow (Fig. 12b); it is often well jointed and displays columnar jointing (sometimes multi-tiered). In the localities that we have studied (mentioned earlier) the thickness can be up to 20 m. In the northern parts of the province (Malwa Traps, and east of Dhule), where much greater thickness of simple flows (>50 m) have been observed, the central zone is proportionately greater in thickness. Our study has not revealed the presence of segregation structures in the central zones of simple flows. The frequency of vesicles is much lower than that in the crustal zone. In several flows, elongated or flattened silica-filled vesicles have been observed near the interface of the crustal zone and the central zone (Fig. 11). Simple flows in the DVP generally can exhibit well-developed jointing patterns (Fig. 13). These patterns are generally persistent laterally and are referred to as colonnade and entablature (Tomkieff 1940; Long and Wood 1986). Entablature refers to 0.30- to 0.50-m- thick, irregular columns that occur in the upper portions of flows. They commonly deviate from a

Fig. 12. a Photo of flow-top breccia exposed close to Kurundwad (Fig. 2). White amygdales are visible, while individual fragments and horizontal joints filled with calcium carbonate have been marked for clarity. b Central zone (~25 m) of a thick flow exposed in a quarry close to Kurundwad (Fig. 2). Note the relatively welldeveloped columnar jointing and the considerable extent of the flow in this exposure. c Basal zone of a flow exposed in the Dhadgaon area (Fig. 2). Note the smooth nature of the base. X refers to a zone of closely spaced joints, while Y indicates dark glassy bands

perpendicular orientation and may form radiating patterns. In certain flows, two or more levels of entablature are seen, usually separated by platy joints (Fig. 13b, c). Well-developed colonnade structures are usually seen in

41 Fig. 13. a Field sketch of a simple flow exposed in Dive Ghat (Fig. 3a). A refers to the entablature, a zone of randomly oriented jointing that imparts a chaotic appearance to the zone. B refers to the colonnade, characterized by more regular columnar joints. b Sketch of part of a simple flow exposed in Katraj Ghat (Fig. 3a) displaying three tiers of joints, each separated by platy joints. c Log of a simple flow from Katraj Ghat (Fig. 3a) showing multi-tiered columnar jointing. CN is colonnade, EN is entablature, PJ is platy joints, while BZ refers to the basal zone

the lower portions of most simple flows. Columns are commonly perpendicular to the base of the flow and their width ranges from 0.50 to 2.0 m. An upper colonnade zone is also present in some flows. The entablature is often glassy and sometimes shows horizontal vesicular zones. In other flows, no clear distinction between the colonnade and entablature can be made, and the entire flow shows highly irregular jointing and a fine texture. Good examples of entablature and colonnade structures are seen at Nanand-Bhavjai Ghat, Toranmal, etc. (see also De 1996) and the southern parts of the Western Ghats (Katraj Ghat, Dive Ghat, Mahabaleshwar, etc.). The columns often fan out in a radial manner, encircling breccia. The proportion of flows displaying well-developed columnar jointing is, however, much less than that in the CRB. Basal zone Most flows display a sharp, slightly undulating, glassy contact with the top of the underlying flow. This zone has a variable thickness and often shows close spaced jointing and a bluish-black sheen as seen in (Fig. 12c). It is rarely red as in the pahoehoe flows. This is particularly well observed in Dive Ghat and around Kurundwad. Plagioclase phenocrysts can often be observed in this zone. Although this zone can be highly vesicular, pipe vesicles have not been observed. The simple flows rest on ground slopes less than 4 and satisfy the criteria for the development of pipe vesicles (Walker 1987). Hence, the absence of pipe vesicles must be related to the rheology of

the lavas and their emplacement style. Thin patches of breccia are sometimes present at the base, but these are usually not extensive. Sometimes it is very difficult to ascertain whether the breccia at the base of these flows is the flow-bottom breccia of the upper flow, or the flow-top breccia of the lower flow. It is worth noting that no evidence of scouring of the base has been observed in any of the flows (e.g., Fig. 12c).

Emplacement of the flows


Compound pahoehoe flows The compound pahoehoe flows in the DVP display excellent, unambiguous evidence of endogenous growth. In this sense, they are very similar to inflated pahoehoe flows in Hawaii (Hon et al. 1994), the Columbia River Basalt Province (Thordarson and Self 1998), and Australia (Whitehead and Stephenson 1998). The features indicating endogenous growth are (1) compound nature of the flows with sheet lobes displaying thick crusts and vesicle layering. The rapid downward decrease in vesicularity accompanied by an increase in vesicle size observed in the Deccan is a characteristic feature of inflated lavas (Cashman and Kauahikaua 1997);( 2) an abundance of local inflation features such as tumuli. The great thickness and abundance of compound pahoehoe flows in the western DVP strongly suggests a slow, rather than rapid emplacement during sustained eruptive episodes. The preliminary vesicle distribution studies indicate that two neighboring lobes may display different

42

patterns. While more studies are needed to comment on this, it suggests that subtle variations in volatile content, degree of endogenous growth or pulses of inflation, rate of cooling, formation of break-outs, etc., may all control vesicle distribution in pahoehoe lobes. The Deccan Trap flows extend for tens to hundreds of kilometers and the area occupied by these compound pahoehoe flows is almost half of the total area of the province (Fig. 1). However, it is not clear if the presence of inflation features is alone sufficient to explain the great areal extent of flows from the DVP, particularly in the absence of well-developed tube systems. Flows from the CRB are constituted of some thick and extensive (kilometer scale) sheet lobes (Keszthelyi et al. 1999), while the longest flow lobes in the DVP are generally observed to extend over a few hundred meters. There is a predominance of smaller lobes over larger ones (Table 3). It is doubtful if such small-scale inflation features would be sufficient to enable the long distance transfer of lava. The large areal extent of flows in such a scenario can theoretically be explained by: 1. The efficient delivery of lava through well-developed tube systems. The efficacy of lava tubes in transferring lava over long distances and in minimizing cooling is well known and has been discussed in detail by Keszthelyi (1995); Keszthelyi and Self (1998); Kauahikaua et al. (1998), etc. Recently, some sinuous lava tubes have been reported by Misra (2002) in parts of the western DVP. Evidence for a small tube system has also been documented by Duraiswami et al. (2001), based on the disposition of tumuli. If these features are indeed tubes, they could shed considerable light on the emplacement of the flows. However, in a large part of the DVP, tubes (sensu stricto) have not been discovered. Tubes in areally extensive systems could also be more like sheets in geometry (R.A.F Cas, personal communication). Jim Kauahikaua (personal communication) has pointed out that the term tube need not be used in a restrictive sense and that other features such as elongate tumuli could also play the same role (e.g., Kauahikaua et al. 1998). The question in the DVP, however, is whether elongate tumuli could transfer lava over the distances that the flows are presently seen to have flowed. This is further elaborated upon in the next point. 2. The long distance transfer of lava via sheet flows: Keszthelyi and Self (1998) discuss the efficacy of sheet flows in transporting lava over long distances. Their modeling suggests that a sheet with a 5-m-thick crust emplaced on a slope of 1% with an effusion rate of 72 m3/s will, in theory, lead to conditions of zero cooling. However, in this case, a single sheet greater than 100 km long may be expected. So far, there is very sparse evidence of the existence of such features in CFB provinces including the CRB. On the contrary, the longest lobes in the CRB are of the order of a few kilometers, while in the DVP, they are even smaller. Could the transfer have been achieved by a system of

interconnected and overlapping lobes, now anastomosed, rather than well-defined tubes? In such a scenario, the flow field might advance by the constant budding of new pahoehoe sheets at the fronts or tops of previously inflated ones, and by ephemeral vent formation. Some evidence in favor of this mechanism may be obtained from the ubiquity of large, dyke-like squeeze-ups that were described earlier (see also Karmarkar 1978). The important question, though, is whether these conduits could be thermally efficient to continue to feed lava from one lobe to the other over long periods of time and over large distances. 3. The assumption that several emplacement units were fed simultaneously from different fissure segments of long vent systems (Keszthelyi et al. 1999): this implies that the eruptions were truly polycentric and that lava extruded out of widely scattered vents throughout the province as proposed by Marathe et al. (1981) and Kale et al. (1992). However, concrete evidence for this is presently lacking. Simple flows The thick, laterally extensive simple flows in the southern and eastern parts of the Deccan have been considered to be aa flows by previous workers (e.g., Godbole et al. 1996). A study of these flows in the western DVP indicates that most of these flows do not satisfy the criteria (e.g., Macdonald 1972; Peterson and Tilling 1980) for classifying them as aa flows sensu stricto, especially proximal-type aa (Rowland and Walker 1987). They do not show the presence of flow-bottom breccia (although in rare cases it has been observed) or irregular, twisted vesicles that characterize true aa flows. Similarly, entrains of the flow-top clinker are often found within the massive cores of aa flows, a feature which is not observed in the DVP. A strong piece of evidence against their being aa flows is their broad (at least a few km) sheet-like morphology and the absence of levees and other channelized flow-related features. This is quite unlike aa flows, which tend to have a narrow, elongate morphology owing to rapid emplacement in open channels (see Keszthelyi et al. 1999). The flow tops of the simple flows in the DVP contain a high frequency of small, almost spherical vesicles and are hence more akin to disrupted pahoehoe crusts. The appearance of the FTB is very different from the jagged and spinose clinker that characterizes aa flows. There is a possibility that the simple flows thickened initially by inflation and developed the flow-top breccias later (Duraiswami et al. 2003). The brecciation may have been initiated as a response to an increase in flow rate or viscosity during the later stages of emplacement. In other words, these flows are somewhat similar to slabby pahoehoe flows/pahoehoe flows with rubbly tops. However, features indicative of inflation are rarely observed. Each simple flow appears to have been emplaced as a broad sheet over a low but consistent gradient during a

43

single eruptive pulse. The volume erupted during each pulse as well as the volumetric flow rate was sufficient to produce a flow of considerable lateral extent. This probably prevented the break down of the flow/flow front into discrete flow lobes. The presence of flow-top breccias and elongated vesicles in the cores of these flows suggests the influence of substantial shearing during emplacement, as well as relatively higher flow rates. However, neither channelized flow nor scouring of the substrate seems to have occurred. It is improbable that the simple flows were emplaced in a manner analogous to typical aa flows, or as turbulent sheets at extremely high discharge rates as proposed by Shaw and Swanson (1970). Duraiswami et al. (2003) explore the possibility of these flows being akin to either inflated pahoehoe sheet flows, or distal type aa flows. They discuss the possible impact of plume-generated uplift and steepening gradients to explain the large areal extent of simple flows. While certain simple flows may have been emplaced as distal type aa, many of them are sufficiently distinct in morphology from true aa and true pahoehoe flows, and warrant a more systematic and detailed study. It is likely that these flows originated as relatively fluid lava with low or no yield strength. The highly vesicular nature of the crust, whether preserved or disrupted, suggests that these lavas were rich in dissolved volatiles. The effect of volatiles on the viscosity of lava is not very well constrained (Keszthelyi and Self 1998). In a dissolved state, volatiles may decrease the viscosity of the lava (Sparks and Pinkerton 1978). Exsolution of volatiles and the presence of bubbles will generally lead to an increase in viscosity since the bubbles act as rigid spheres (Keszthelyi and Self 1998). However, at high strain rates the bubbles may be sheared and this will actually lower the viscosity (op cit). The high initial fluidity of the simple flows may have been a result of the presence of a high proportion of dissolved gases, and even after exsolution, the strain rates were probably high enough to maintain the pre-exsolution viscosity. Evidence for high strain rate comes from the ubiquitous FTB. The fact that fragments in the FTB are often highly vesicular suggests that substantial degassing preceded the brecciation. However, stretched vesicles indicative of such strain rates are not observed in the FTB, but are observed in the central zones of such flows. It is now imperative to ascertain whether the present thickness of these flows is the same as that when they were erupted, or is a result of some process of endogenous growth. Similarly, it is important to determine the duration of emplacement for a single simple flow. Unfortunately, since no quantification of morphology of these flows exists, and since they have poorly preserved crusts, the equation of Hon et al. (1994) cannot be directly used to calculate the duration. The morphology of the simple flows in the DVP needs to be studied in greater detail since they may prove to be crucial in answering questions regarding flow rates, and mechanisms of emplacement of CFB lavas. Macdonald (1972) has described similar flows from the Yakima basalts of the CRB. He mentions that many

Yakima basalt flows do not fit the descriptions of typical aa flows, neither are they identical to pahoehoe flows. These flows possess flow-top breccias similar to those of the simple flows, which constitute 2035% of the flow thickness. Other characters of the Yakima basalt flows as described in Macdonald (1972) are also similar to those of the simple flows in the DVP. It is likely that the two are a product of a similar emplacement process. Some of the columnar flows from the DVP may have been emplaced in pre-existing topographic lows/river valleys. The localized ponding probably led to the development of multi-tiered columnar jointing with glassy entablatures and fanning columns. According to De (1996), the presence of entablature is probably a result of the near vertical nature of isotherms during the later stages of cooling of lava. Such flows are not as common in the DVP as compared to the CRB, where they seem to be quite common (Long and Wood 1986). In the DVP, they occur in the northern fringes of the province, close to an ancient zone of weakness. It is quite possible that the paleotopography was favorable for the ponding of flows. However, this may not necessarily be the only factor responsible. Thick flows with multi-tiered columnar jointing commonly occur higher up in the stratigraphic sequence, which means that they cannot have been ponded in the pre-eruptive topographic depressions. Although the duration of hiatuses between the emplacement of two successive flows are not known, the absence of intervening sediments and/erosion suggests that these may not have been substantial enough to allow significant topography to develop. Hence, the origin of the multitiered jointing may need to be explained by factors other than ponding and damming of drainage.

Implications for the emplacement of flood basalt lavas


The observations of flows from the DVP emphasize the need for physical volcanological studies in CFB provinces other than the CRB. It is probable that such studies will reveal the similarities as well as differences in the styles of eruption of flows from different provinces. In a recent study, Anderson et al. (1999) observe that CFB flows typically display spectacular columnar jointing, as opposed to Hawaiian inflated flows that have crudely columnar tops and massive interiors. They use this observation to suggest that the emplacement of flood basalts is fundamentally different from that of Hawaiian pahoehoe flows. Self et al. (2000), in their discussion of the pulsed inflation model of Anderson et al. (1999), draw a distinction between hummocky pahoehoe flows as found in Hawaii and large pahoehoe sheet flows/lobes seen in the CRB. They argue that CFB provinces are dominantly constituted of sheet flows, and hence conclusions derived from hummocky flows cannot be directly applied to the emplacement of CFB lavas. The present study clearly highlights the fact that the CRB is not representative of all CFB provinces, as has

44

also been mentioned by Long and Wood (1986) in their discussion of multi-tiered jointing. Previous studies (Bondre et al. 2000; Duraiswami et al. 2001, 2002, 2003), as well as the present study suggest that flows very similar to those in Hawaii are abundant in the DVP. Small flow lobes and tumuli are as common as, if not more abundant than sheet flows. Models of pahoehoe lava emplacement in the CRB (e.g., Thordarson and Self 1998) deal with extensive, thick inflated sheets (>25 m thick). It remains to be investigated whether such models may be directly applicable to the strongly compound pahoehoe flows from the DVP. At the same time, simple flows from the DVP are similar to certain flows from the CRB. It thus seems likely that there were considerable variations in the styles of emplacement of flows from different CFB provinces, and even within a single province as is exemplified by the DVP. It is therefore critical to insure that any general model of flood basalt emplacement takes all CFB provinces into account; the possibility remains, however, that there may not really be any such general model (J. Kauahikaua, personal communication). At the same time, it is very important to document these variations and model their emplacement. These variations may ultimately reflect upon the differences in the tectonics of formation of the CFB provinces, as has been discussed by Duraiswami et al. (2003). As compared to other provinces, studies of flow morphology and quantitative studies of lava flow emplacement are lagging far behind in the DVP. The present study is only a small attempt to review the existing state of knowledge about flows from this province, and needs to be followed up with rigorous quantification.
Acknowledgements We have benefited greatly from our correspondence with Profs. G.P.L. Walker, Stephen Self, Laszlo Keszthelyi and Jon Stephenson. We thank Dr. J. Kauhikaua for his encouraging comments, and Dr. R.A.F Cas and Dr. Tim Druitt for their critical and careful reviews. Discussions with Dr. Vivek Kale and Shreyas Mangave were highly illuminating. Ninad Bondre and Gauri Dole are grateful to Prof. K.V. Subbarao for giving them an opportunity to participate in the Penrose Deccan 2000 field conference. Ninad Bondre would also like to thank his parents and Zu Watanabe for their encouragement.

References
Anderson SW, Stofan ER, Smrekar SE, Guest JE, Wood B (1999) Pulsed inflation of pahoehoe lava flows: implications for flood basalt emplacement. Earth Planet Sci Lett 168:718 Anderson SW, Stofan ER, Smrekar SE, Guest JE, Wood B (2000) Reply to: Self et al. discussion of Pulsed inflation of pahoehoe lava flows: implications for flood basalt emplacement. Earth Planet Sci Lett 179:425428 Aubele JC, Crumpler LS, Elston WE (1988) Vesicle zonation and vertical structure of basalt flows. J Volcanol Geotherm Res 35:349374 Auden J B (1949) Dykes in western India. Trans Natl Inst Sci, India 3:123157 Beane JE, Turner CA, Hooper PR, Subbarao KV, Walsh JN (1986) Stratigraphy, composition and form of the Deccan basalts, Western Ghats, India. Bull Volcanol 48:6183 Bhattacharjee S, Chatterjee N, Wampler JM (1996) Timing of the NarmadaTapti rift reactivation and Deccan volcanism:

geochronological and geochemical evidence. Gondwana Geol Mag Spl 2:329340 Bondre NR, Dole G, Phadnis VM, Duraiswami RA, Kale VS (2000) Inflated pahoehoe lavas from the Sangamner area of the western Deccan Volcanic Province. Curr Sci 78:10041007 Cashman KV, Kauahikaua JP (1997) Re-evaluation of vesicle distributions in basaltic lava flows. Geology 25:419422 Courtillot VE, Bease J, Vandamme D, Montigny R, Jaeger JJ, Cappetta H (1986). The Deccan flood basalts at the CretaceousTertiary boundary? Earth Planet Sci Lett 80:361374 Cox KG and Hawkeshworth C J (1985) Geochemical stratigraphy of the Deccan Traps at Mahabaleshwar, Western Ghats, India, with implications for open system magmatic processes. J Petrol 26:355377 De A (1996) Entablature structure in the Deccan Trap flows: its nature and probable mode of origin: Gondwana Geol Mag Spl 2:439447 Deshmukh SS (1988) Petrographic variations in compound flows of Deccan Traps and their significance. In: Subbarao, KV (ed) Deccan flood basalts. Mem Geol Soc India 10:305319 Devey CW, Lightfoot PC (1986) Volcanological and tectonic control of stratigraphy and structure in the western Deccan traps. Bull Volcanol 48:195207 Dole G, Bondre, N, Duraiswami R A, Kale V S (2002) Discussion on Arterial system of lava tubes and channels within Deccan volcanics of western India by K S Misra. J Geol Soc India 60:597599 Duraiswami RA, Bondre NR, Dole G, Phadnis VM, Kale VS (2001) Tumuli and associated features from the western Deccan volcanic province, India. Bull Volcanol 63:435442 Duraiswami RA, Bondre N, Dole G (2002) Morphology and structure of flow-lobe tumuli from Pune and Dhule areas, western Deccan Volcanic Province. J Geol Soc India 60:5765 Duraiswami RA, Dole G, Bondre N (2003) Slabby pahoehoe from the western Deccan Volcanic Province: evidence for incipient pahoehoe-aa transitions. J Volcanol Geotherm Res 121:195 217 Godbole SM, Rana RS, Natu SR (1996) Lava stratigraphy of Deccan basalts of western Maharashtra. Gondwana Geol Mag Spl 2:125134 Goff F (1996) Vesicle cylinders in vapour differentiated basalt flows. J Volcanol Geotherm Res 71:167185 Hon K, Kauahikaua J, Denlinger R, Mackay K (1994) Emplacement and inflation of pahoehoe sheet flows: observations and measurements of active lava flows on Kilauea Volcano, Hawaii. Geol Soc Am Bull 106:351370 Kaila KL (1988) Mapping the thickness of Deccan Trap flows in India from DSS studies and inferences about a hidden Mesozoic basin in the Narmada-Tapti region. In: Subbarao KV (ed) Deccan flood basalts. Mem Geol Soc India 10:91116 Kale VS, Kulkarni HC, Peshwa VV (1992) Discussion on a geological map of the southern Deccan Traps, India and its structural implications. J Geol Soc Lond 149:473478 Karmarkar BM (1978) The Deccan Trap basalt flows of the Bor Ghat section of Central Railway. J Geol Soc India 19:106114 Kauahikaua J, Cashman, KV, Heliker C, Hon KA, Mangan MT, Mattox TN, Thornber CR (1998) Observations on basaltic lava streams in tubes from Kilauea Volcanic, Island of Hawaii. J Geophys Res 103:2730327323 Keszthelyi L (1995) A preliminary thermal budget for lava tubes on the earth and planets. J Geophys Res 100:2041120420 Keszthelyi L, Self S (1998) Some physical requirements for the emplacement of long basaltic lava flows. J Geophys Res 103:2744727464 Keszthelyi L, Self S, Thordarson T (1999) Application of recent studies on the emplacement of basaltic lava flows to the Deccan Traps. In: Subbarao, KV (ed) Deccan flood basalts. Mem Geol Soc India 10:485520 Long PE, Wood BJ (1986) Structures, textures and cooling histories of Columbia River basalt flows: Geol Soc Am Bull 97:1144 1155

45 Macdonald GA (1972) Volcanoes. Prentice Hall, Englewood Cliffs, p 90 Mahoney JJ, Sheth HC, Chandrasekharam D, Peng ZX (2000) Geochemistry of flood basalts of the Toranmal section, northern Deccan Traps, India: implications for regional Deccan stratigraphy. J Petrol 41:10991120 Marathe SS, Kulkarni SR, Karmarkar BM, Gupte RB (1981) Variation in the Deccan Trap volcanicity of western Maharashtra in time and space. Mem Geol Soc India 3:143152 Misra KS (2002) Arterial system of lava tubes and channels within Deccan volcanics of western India. J Geol Soc India 59:115 124 Mitchell C, Widdowson M (1991) A geological map of the southern Deccan Traps, India and its structural implications. J Geol Soc Lond 148:495505 Morgan JW (1972) Plate motions and deep mantle convection. Geol Soc Am Bull 132:722 Nichols RL (1936) Flow-units in basalt. J Geol 44:617630 Peterson DW, Tilling, RI (1980) Transition of basaltic lava from pahoehoe to aa, Kilauea volcano, Hawaii: field observations and key factors. J Volcanol Geotherm Res 7:271293 Powar KB (1987) Evolution of the Deccan Volcanic Province. In: Presidential Address, 74th Indian Science Congress, Bangalore, 30 pp Rajarao CS, Sahasrabuddhe YS, Deshmukh SS, Raman R (1978) Distribution, structure and petrography of the Deccan Traps, India. Excerpts recently published in: Subbarao KV (ed) Deccan Volcanic Province. Mem Geol Soc India 43 1:401414 Rowland SK, Walker GPL (1987) Toothpaste lava: characteristics and origin of a lava structural type transtional between pahoehoe and aa. Bull Vocanol 49:631641 Self S, Keszthelyi L, Thordarson T (2000) Discussion of: Pulsed inflation of pahoehoe lava flows: implications for flood basalt emplacement, by Anderson et al. Earth Planet Sci Lett 179:421423 Shaw HR, Swanson DA (1970) Eruption and flow rates of flood basalts. Proceedings of the 2nd Columbia River Basalt Symposium, East Washington State College Press, Cheney, pp 271 299 Solanki JN, Bhattacharya DD, Jain AK, Mukherhee A (1996) Stratigraphy and tectonics of the Deccan Traps of Mandla. Gondwana Geol Mag Spl 2:101114 Sparks RSJ, Pinkerton H (1978) Effect of degassing on rheology of basaltic lava. Nature 276:358386 Subbarao KV (ed) (1988) Deccan flood basalts. Bangalore, India. Mem Geol Soc India 10:393 pp Subbarao KV (ed) (1999) Deccan Volcanic Province. Bangalore, India. Mem Geol Soc India 43:547 pp Subbarao KV, Hooper PR (1988) Reconnaissance map of the Deccan Basalt Group in the Western Ghats. India. In: Subbarao KV (ed) Deccan flood basalts. Mem Geol Soc India 10: (enclosure) Subbarao KV, Bodas MS, Hooper PR, Walsh, JN (1988) Petrogenesis of Jawhar and Igatpuri Formations of western Deccan Trap Province. In: Subbarao KV (ed) Deccan flood basalts. Mem Geol Soc India 10:253280 Thorat PK (1996) Occurrence of lava channels and tubes in the western part of Deccan Volcanic Province. Gondwana Geol Mag Spl 2:449456 Thordarson T, Self S (1998) The Roza Member, Columbia River Basalt Group: a gigantic pahoehoe lava flow field formed by endogenous processes? J Geophys Res 103:27,41127,445 Tomkieff SI (1940) The basalt lavas of the Giants Causeway District of Northern Ireland. Bull Volcanol 6:89143 Walker GPL (1969) Some observations and interpretations on the Deccan Traps. Recently published in: Subbarao KV (Ed) Deccan Volcanic Province: Mem Geol Soc India 43 1:367395 Walker GPL (1971) Compound and simple lava flows and flood basalts. Bull Volcanol 35:579590 Walker GPL (1987) Pipe vesicles in Hawaiian basalt lavas: their origin and potential as paleoslope indicators. Geology 15:8487 Walker GPL (1993) Basaltic-volcano systems. In : Pritchard, HM, Alabaster, T, Harris, NBW, Neary, CR (eds) Magmatic processes and plate tectonics. Geol Soc Spec Publ 76:338 Watts AB, Cox KG (1989) The Deccan Traps: an interpretation in terms of progressive lithospheric flexure in response to a migrating load. Earth Planet Sci Lett 93:8597 West WD (1959) The source of Deccan Trap flows. J Geol Soc India 1:4452 Whitehead PW, Stephenson PJ (1998) Lava rise ridges of the Toomba basalt flow, north Queensland, Australia. J Geophy Res 103:27,37127,382 Wilmoth RA, Walker GPL (1993) P-type and S-type pahoehoe: a study of vesicle distribution patterns in Hawaiian lava flows. J Volcanol Geotherm Res 55:129142

You might also like