The association of alkaline igneous rocks and carbonatites (ARCs) with intracontinental rifts has long been recognized. Deformed alkaline rocks (such as nepheline syenites) and carbonatites (DARCs), which form a small subset (<10%) of ARCs and carbonatites, have become a focus of attention since it was recognized that in some places they are associated with suture zones marking locations where oceans have closed. The association of DARCs with sutures can be readily understood in terms of the operation of the Wilson cycle in the opening and the closing of ocean basins. The Wilson cycle hypothesis for the origin of DARCs has been tested for the Proterozoic of both Africa and India. For Africa, an additional hypothesis that we have also tested is that ARCs in intracontinental rifts represent magmas made by partial melting of DARC material in the underlying mantle lithosphere. A limitation of the African and Indian studies is that they are local and regional. Only a worldwide geoinformatic study can fully test the two hypotheses. If one of the hypothesis can be invalidated that will have profound petrogenetic significance. In this study, we describe our ongoing geoinformatic work on the worldwide distribution of ARCs and DARCs, including the databases, techniques, and operating procedures. To illustrate our approach, we report on a local analysis that we have carried out on the ARCs and DARCs of the Kola Peninsula in Arctic Russia.

The methods of geoinformatic research can be used to address fundamental questions in the earth sciences. Here, we show the initial application of those methods, using alkaline igneous rocks and carbonatites, to basic questions in (1) igneous petrology and (2) tectonics.

Igneous Petrology

There are currently two hypotheses about the origin of alkaline igneous rocks, such as nepheline syenites and carbonatites: one hypothesis (see, for example, Bell and Tilton, 2002) suggests a source in the convecting mantle (perhaps a plume) (Fig. 1A), and the other hypothesis indicates that (see, for example, Burke et al., 2003) alkaline rocks and carbonatites are derived from partial melting of a part of the underlying mantle lithosphere that includes deformed alkaline rock and carbonatite (Fig. 1B). It is not appropriate in this paper, which focuses on geo-informatics, to get too far into a discussion of the broad petrogenetic implications of the origin of alkaline igneous rocks and carbonatites (ARCs); we confine ourselves to pointing out the broad petrogenetic significance of whether ARCs originate in the mantle lithosphere or in the convecting mantle. If those rocks come from the convecting mantle, there is an implication that reservoirs with distinct chemical and isotopic compositions have remained isolated within the convecting mantle perhaps for periods as long as 3 b.y. If ARCs are derived from source material within the mantle lithosphere, the question arises: In what environment did the first ARC source materials form? Appleyard (1974) suggested that an evaporite source was possible, but as far as we know, the issue has not been revisited in recent years.


There is continuing debate about the extent to which plate-tectonic processes were involved in the evolution of the continental crust and lithosphere during Precambrian times. Where Precambrian suture zones can be identified, marking places in which the Wilson cycle has operated, there is strong evidence that oceans were opening and closing during the evolution of the Precambrian continental lithosphere.

The association of ARCs (nepheline syenites, their volcanic equivalents, tephrites and phonolites, and carbonatites) with intracontinental rift systems has been well established (Bailey, 1974, 1977, 1992) and has lately been confirmed for large parts of Earth in the catalogs of Woolley and his collaborators (Woolley, 1987, 2001, 2001; Kogarko et al., 1995). A new focus is on the deformed alkaline rocks and carbonatites (DARCs) showing gneissic structure, which form <10% of ARCs. ARCs and DARCs are identical in chemistry and petrology but contrast greatly in structure. ARCs form intrusions, usually round and often clearly subvolcanic, but DARCs form concordant lenses often among gneisses and granulites.

Operation of the Wilson cycle has been invoked to reconcile the compositional identity of ARCs and DARCs with their contrasting structure (Fig. 2; Burke et al., 2003). ARCs are interpreted, as is conventional, to have been erupted into intracontinental rifts (Fig. 2A), and DARCs are considered to represent ARCs that have been involved in ocean opening and closing. In the process of ocean opening, ARCs come to lie on a rifted continental margin as a young ocean develops from an intracontinental rift (Fig. 2B). No great changes happen to ARC rocks during ocean opening and as the ocean begins to close (Fig. 2B), but in the final stage of the Wilson cycle, when an island arc or a continent collides with a rifted continental margin, the ARC becomes deformed into a DARC (Fig. 2C). The hypothesis of Burke et al. (2003) predicts that DARCs should be found on suture zones within continents. Those suture zones mark sites of both a rifted continental margin and a later continental collision at that margin.

If DARCs lie on suture zones, then the concentration of ARCs in intracontinental rifts can also be explained. Wilson (1966) showed that rifts commonly develop above existing sutures. ARCs in intracontinental rifts can therefore be interpreted to represent material derived from melting that involved previously formed DARC rocks at a suture zone within the underlying mantle lithosphere (Figs. 1B and 2E).

Testing the Wilson cycle model of ARC and DARC origins has just begun. It has been found to work well in Africa (Fig. 3; Burke et al., 2003) and in India (Fig. 4; Leelanandam et al., 2005). Work on testing the model is also beginning in the Canadian cordillera (Johnston et al., 2003). These studies must be regarded as preliminary, because the model can only be fully tested in a global geoinformatic study. Here, we outline the structure of the geoinformatic work, which we are beginning to address this problem. We include a description of what databases are now available and how we plan to improve them. We illustrate the kinds of results that can be hoped for with an example from the Kola Peninsula of Arctic Russia.


We use three main existing databases. During the course of the study, those databases will be checked, refined, augmented, and modified. The databases are: the Woolley catalogs of alkaline rocks and carbonatites, the world map of rifts, and the world map of sutures.

The Woolley Catalogs of Alkaline Rocks and Carbonatites

The content and structure of the three volumes of this important set of catalogs are described in the following.

Volume 1: Alkaline Rocks and Carbonatites of the World Part 1: North and South America (1987), by Alan R. Woolley; publisher: University of Texas Press (Austin) and British Museum (Natural History), 216 large-format (14 inch × 10 inch) pages.

Complexes discussed are from nineteen countries: Canada, Greenland, Mexico, and the United States (North America); Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, Guyana, Haiti, Honduras, Paraguay, Peru, Uruguay, and Venezuela (Central and South America). Discussions of individual alkaline rock occurrences vary greatly but commonly include place names, latitudes and longitudes, petrology, mineralogy, ages (mainly isotopic ages), geological maps of varied detail and scale (redrafted for the catalog), and in some cases economic notes and geophysical information. Canada is one of the most fully treated countries, with 165 localities and ∼270 references; a two-page map, showing the locations of all the localities discussed, and 81 localities and regional maps are provided. The United States, another fully treated country, has 144 localities and 67 maps. Greenland has 144 localities and 29 maps. By contrast, for several countries, only one occurrence (e.g., Honduras) to three occurrences (e.g., Uruguay) were cataloged, and there are no detailed maps.

Volume 2: Alkaline Rocks of the World Part 2: Former Soviet Union (1995), by L.N. Kogarko, V.A. Kononova, M.P. Orlova, and A.R. Woolley; publisher: Chapman and Hall, London, 226 pages. The format and coverage is similar to that of Volume 1. Occurrences are described by province; 23 provinces are distinguished and over 400 individual occurrences are described. There are 253 maps, including one map of the whole region, 23 maps showing the provinces, and 229 maps showing individual occurrences. References are numerous; for example, there are over 200 for the Kola-Karelia region. English-language references are distinguished.

Volume 3: Alkaline Rocks and Carbonatites of the World Part 3: Africa (2001), by A. Woolley; publisher: Geological Society (London), 372 pages. The structure of Volume 3 is similar to that of the two previous volumes; 842 alkaline igneous rock complexes are described from occurrences in 40 countries, and 348 maps illustrate those descriptions. About 4% of the occurrences (32) are DARCs, and ∼90% of those DARCs (28) lie within known suture zones (Burke et al., 2003).

World Map of Rifts.

Rifts are defined as elongate depressions overlying places where the lithosphere has ruptured in extension. In the interior of plates, away from plate boundaries, the lithosphere is generally in compression (Zoback, 1992). Rifts may form anywhere that extension has locally modified the intraplate stress field. Intracontinental rifts have formed in a variety of tectonic environments (see, for example, Şengör, 1995, his Figure 2.5). Several maps showing the worldwide distribution of intracontinental rifts have been compiled over the past 30 yr (Burke et al., 1977; Şengör, 1995, his Figure 2.2). We use the most up-to-date rift catalog (that of Şengör and Natalin, 2001; Fig. 5) to test the idea that ARCs have been strongly associated with intracontinental rifts during the past 2.5 b.y.

World Map of Sutures.

We use the world map of sutures of Burke et al. (1977), which are clearly in need of updating. For example, the Early Proterozoic suture of the Kola Peninsula (Daly et al., 2001) that we refer to in this paper had not been recognized in 1977.


We utilized geoinformatic capabilities for spatial resolution, manipulation, and dissemination of diverse data categories. For selected data sets, we created a geodatabase, which includes structural data and a regional geological map in shape-file format. Our geodatabase is an Environmental Systems Research Institute (ESRI) relational database system that stores small-size geographic data (<2 GB) and acts as a container for storing spatial and attribute data and the relationships that exist among them. In geodata-bases, a vector data format for features and their associated attributes can be structured to work together as an integrated system using rules, relationships, and topological associations. Also, we included remote-sensing and digital elevation data for selected sites as raster images.

Our longer-term plan is to transfer current data and compile more data in a Microsoft SQL Server 2000 version relational database management system (RDBMS) and make this database accessible to the community through the Internet (pending availability of findings). The database system will be integrated with ArcIMS (ARC Internet Map Server) via ArcSDE (ARC Spatial Database Engine). ArcSDE is a geographic information system (GIS) gateway that facilitates management of spatial data in relational database management systems (SQL server in our case), whereas ArcIMS is currently widely utilized in the integration of local GIS data sources with Internet data sources for display, query, and analysis using a Web browser (see Fig. 6). The database will include digital map data sets using the Woolley catalogs (Woolley, 2001, 1987; Kogarko et al., 1995). We will first compile relevant existing geologic map data in various formats (hard copy, digital maps, open-file reports, etc.). Maps will be reprojected to a common standard projection. Geologic map objects will include polygons (e.g., outcrops) and lines (e.g., faults, unconformities). Polygons and lines will have various attributes, for example, polygon attributes include lithology, age, thickness, and unit name, and line attributes include feature types, orientation, and name. We will generate relevant metadata files that are compatible with metadata standards of the Federal Geographic Data Committee (FGDC) and National Geologic Map Database (NGMDB) (Lehnert et al., 2000; Ryburn, 1999).We will continue incorporating map data for areas that have not yet been covered in the catalogs. Rift and suture maps will be digitized, and the distribution of ARCs and DARCs will be plotted on them.


The “Former Soviet Union” volume of the Woolley catalog (Kogarko et al., 1995) contains descriptions of ARCs and DARCs of the Kola Peninsula that provide an opportunity for a pilot geoinformatic study.

In step 1, we located ∼30 ARCs and 2 DARCs in Kola from the catalog (Kogarko et al., 1995), which gives latitudes and longitudes as well as maps of individual occurrences (Fig. 7). In step 2, a test area around the two main ARCs of Kola (Lovozero and Khibina) was selected. In step 3, a Landsat™ image of the area was acquired. In step 4, maps of the two ARCs and two much smaller DARCs were scanned from the catalog. In step 5, the Landsat image and the scanned maps were projected in the same projection and were imported to ArcGIS and plotted (Fig. 7). The DARC outcrops lie on either side of the area in which the ARCs outcrop. One DARC, Soustova, was shown in the catalog as foliated. That foliation was found to be concordant with the regional gneissic foliation on the Landsat image. A similar, although less obvious, relationship was discerned for the second DARC. In step 6, a literature search showed: (1) that Early Proterozoic rift facies rocks and an ESE-trending Early Proterozoic suture zone have been identified in the neighborhood of the ARCs and DARCs of Figure 7 (Daly et al., 2001), and (2) that a poorly defined ESE-trending rift structure, the Kola aulacogen, has been identified as occupying the region in which the Paleozoic ARC rocks of Kola erupted (Kogarko et al., 1995).

Interpretation of the Kola Data

Daly et al. (2001) used a variety of isotopic systems to establish an Early Proterozoic history of ocean opening and ocean closing in Kola. They distinguished two Archean continental blocks from an intervening belt of rocks with Early Proterozoic ages. The rocks in that intervening belt represent material generated and accreted within and at the margin of an ancient ocean basin. Deformed rift-environment rocks were distinguished within and at the margins of the Archean blocks (Fig. 8).

The Kola DARCs at Soustova and Kurginskii (see Fig. 8) occur (as the Wilson cycle model requires) in proximity to the rifted margin of the northern (present coordinates) side of the Proterozoic ocean. The DARCs also lie close to the suture zone mapped by Daly et al. (2001) that marks the place where that ocean closed (Fig. 8). Our interpretation is that the DARCs acquired their gneissic structure during the collision that closed the Early Proterozoic ocean.

The two largest ARCs in Kola (Lovozero and Khibina) are shown on Figure 8. They are of Devonian age and were erupted into the ESE-trending rift (Kola aulacogen; Kogarko et al., 1995) that formed in response to the Scandian (Early Devonian) continental collision in northern Norway. All the ARCs of the Kola Devonian population were intruded over a short interval ca. 370 Ma (K. Bell, 2005, personal commun.). The location of these ARCs in the Kola aulacogen, which is a Russian word for a rift, confirms that the rift was of the kind formed in response to a continental collision. The type example is the Baikal rift, which formed in response to the Himalayan collision (Molnar and Tapponnier, 1975). The location of Lovozero and Khibina between the two DARCs (Fig. 8) lends support to the ideas: (1) that ARCs erupt into rifts, and (2) that the magma of those ARCs forms by partial melting (specifically decompression melting) of DARCs in the parts of suture zones in which the rifts have developed within the mantle lithosphere (Fig. 8).

We have outlined the main features of a geo-informatic study, which, when completed, will have accommodated and analyzed the worldwide distribution of ARCs and DARCs. We have also illustrated the application of our methods using the example of the Kola Peninsula of Arctic Russia.

Tests of a Wilson cycle model, which embraces the two hypotheses: (1) DARCs are concentrated in suture zones, and (2) ARCs later erupt into rifts formed on top of those suture zones, have now been carried out for Kola (this paper), Africa (Burke et al., 2003) and India (Leelanandam et al., 2005). Results in these areas: (1) provide evidence in support of the idea that DARCs occupy ancient suture zones and therefore that our understanding of the distribution of DARCs contributes to tectonic understanding by providing evidence of where oceans have opened and closed within continents; (2) provide evidence in support of the idea that the occurrence of ARCs within rifts in close proximity to older DARCs is compatible with the hypothesis that ARCs form from melting involving DARC material in the underlying lithosphere. If it can be resolved whether a mantle lithosphere or a convecting mantle source is involved in the origin of ARCs, it will be a useful contribution to petrogenesis, as well as a contribution to the present, rather confusing discussion of the nature of mantle plumes.

Attribution: You must attribute the work in the manner specified by the author or licensor (but no in any way that suggests that they endorse you or your use of the work).
Noncommercial ‒ you may not use this work for commercial purpose.
No Derivative works ‒ You may not alter, transform, or build upon this work.
Sharing ‒ Individual scientists are hereby granted permission, without fees or further requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in other subsequent works and to make unlimited photocopies of items in this journal for noncommercial use in classrooms to further education and science.