Kimberlite and lamproite are rare alkaline rocks that exhibit high concentrations of alkalis and volatile contents. These rocks are unique in their geochemical characteristics and are highly susceptible to alteration processes which can result in the depletion of mobile elemental (such as K, Na, and LILEs) concentrations. Despite geochemical similarities, kimberlites and lamproites have certain differences and do not plot within their respective fields when using the extant discrimination diagrams. To address this issue, here we develop new discrimination diagrams based on HFSEs (such as Zr, Nb, Hf, Ta, Ti, Y), Ce and Yb. HFSEs have high affinity to alkaline rocks, and also remain immobile during various magmatic and post-magmatic processes. Although their concentration or geochemical behaviour varies distinctively between lamproites and kimberlites, which make them ideal for a precise classification of kimberlites and lamproites. The newly proposed discrimination diagrams are effective in discriminating kimberlite from lamproite with 92 to 97

Minerals are vital resources that directly influence economic growth and industrial development. Accurate identification and classification of minerals from geological samples are essential for efficient exploration, extraction, and sustainable utilization. Charnockite, a metamorphic rock widely distributed in the Eastern Ghats Mobile Belt, contains a complex mixture of oxides, trace elements, and rare earth elements, making it suitable for machine learning based classification. This study develops a systematic framework to classify the mineral composition of Charnockite rock using data augmentation and Multinomial (SoftMax) Logistic Regression. The workflow begins with preprocessing steps such as handling missing values, detecting outliers, and transforming non-normalized variables into model ready form. To address the limited dataset size, a Dirichlet- based augmentation technique is applied to generate synthetic yet realistic oxide compositions, improving class balance and training robustness. Exploratory Data Analysis (EDA) through correlation matrices, linear and multivariate analyses, and density plots is performed to provide insights into feature distributions and interrelationships. The classification model is then evaluated using accuracy, precision, recall, and F1-score. Results demonstrate an accuracy of 82.85 %, with consistently high precision and recall across classes. Linear and multivariate analyses further reveal redundant as well as higher order dependencies among oxides, supporting dimensionality reduction and enhancing interpret- ability. This study confirms that data augmentation combined with Multinomial Logistic Regression offers a reliable and interpretable solution for mineral classification in geological datasets. The framework can be extended to larger datasets and diverse lithological units, contributing to faster and more sustainable mineral resource exploration.

The Sukinda ultramafic complex in India comprises precisely two areas: Kaliapani (KLPN) and Katpal (KTPL). These areas consist of a sequence of lithotypes, including orthopyroxenite, dunite, serpentinite, and chromitite, displaying a rhythmic layering of rocks. These rocks exhibit a cumulate texture and stand out due to their elevated Mg# (78.43–93.20), Cr (905.40–58,799 ppm), Ni (193.81–2790 ppm), Al2O3/TiO2 (27.01–74.06), and Zr/Hf (39.81–55.24) ratios, while possessing lower TiO2 contents (0.01–0.12 wt%). These ultramafics, characterized by low Ti/V (0.83–19.23) and Ti/Sc (7.14–83.72) ratios, negative anomalies of Zr, Hf, Nb, and Ti in a primitive mantle-normalized spider diagram, indicate that the ultramafics originate from a depleted mantle source. Furthermore, the presence of enriched LREE compared to HREE, a negative Eu anomaly, and enrichment of Th, U, and negative Nb anomalies suggest a subduction setting. The whole-rock geochemical data reveal high levels of MgO, Cr, and Ni, as well as low TiO2 and CaO/Al2O3 ratios and high Al2O3/TiO2 ratios. Moreover, the mineral chemistry data of the ultramafic rocks show high-Mg olivine (Fo 90.9−94.1) in dunite, high-Mg orthopyroxene (En 90.4–90.7) in orthopyroxenite, and high Cr# (0.68–0.82) and low Mg# (0.40–0.54) in chromite, alongside significant Al2O3 (9.93–12.86 wt%) and TiO2 (0.20–0.44 wt%) contents in the melt. Such geochemical characteristics strongly suggest that the Sukinda ultramafic originates from the fractional crystallization of a boninitic parental magma, which is derived from the second-stage melting in a depleted metasomatized mantle source within a supra-subduction zone tectonic setting.

AbstractThis communication reports novel geochemical and geochronological data of granite from the southeastern part of the Bastar Craton, Central India. The studied samples are leucocratic in appearance and composed of quartz, K-feldspar, plagioclase feldspar, and biotite in decreasing order of abundances. Apatite, sphene, and zircon occur as accessory minerals. The SiO2 and Al2O3 content of the studied sample varies between 61 and 69 wt.% and 13 and 15 wt.%, respectively. The alkali oxides, K2O, and Na2O content ranges between 3 and 6 wt.% and 2 and 3 wt. %, respectively. In the primitive mantle normalized spider diagram, the granites exhibit a negative Nb–Ti, Sr anomaly, and a positive Pb–Th anomaly. Similarly, in the REE normalized spider plot, the granites exhibit a strongly fractionated trend La/YbCN=10.90−28.4 with a negative Eu anomaly (0.42-0.70). The zircon saturation in silicate melt yields crystallization temperature (Tzr) ~650 to 800°C for the Eastern Bastar Craton rocks. The P-T pseudosection modeling implies EBC granites which are crystallized at 700-750°C, at 0.4 to 0.6 GPa. The SHRIMP U-Pb ages from magmatic zircon yield an upper intercept at ~2470 Ma and a lower intercept at ~2100 Ma. When combined with the results of P-T pseudosection modeling, the geochemical and geochronological data classifies the Eastern Bastar Craton rocks as A2 granites that were emplaced during the amalgamation of Archean blocks leading to extended Ur formation. The ~2100 Ma age is correlated with mafic dyke emplacement and the Bastar Craton–Yilgarn Craton block disintegration before Paleoproterozoic Columbia supercontinent assembly.

Composition of basalts in magmatic arcs influenced by the subducting lithosphere, mantle wedge, dehydration of oceanic crust, and/or crustal assimilation beneath the arc. In this paper, we compiled earlier published geochemical data of Dongargarh basalts to decipher the genesis of volcanic rocks. SiO 2 vs (FeO + MgO) plot of basalt suggests the volcanic rocks are tholeiitic in composition. Primitive mantle and REE normalized plots indicate either the source was enriched mantle or a possible interaction of depleted magmatic source with the Paleoarchean continental crust in the Bastar Craton. The primitive mantle normalized diagram shows a negative anomaly of Nb, Ti, and Ta indicates subduction-related magmatism. In addition to the basalt composition, variation diagrams for tectonic settings represent the continental arc-related magmatism. From the available geochemical data of basalts and earlier studies on Dongargarh volcanic, there was an oceanic ridge that was subducted beneath the continental plate. The source of Pitepani basalts was significantly enriched in HFSE and REE as compared to mid-oceanic basalts. Thus the study finds the volcanic rocks are part of enriched mantle source that formed in the subduction-related magmatism.

Occurrences of the small, elongated, irregular shape of charnockite patches along the foliation planes of the leptynites in Digapahandi area, Ganjam district, Odisha of Eastern Ghats Belt (EGB) are reported. Petrography and mineral chemistry of charnockite patches and host leptynites suggest that both the rocks have analogues mineral assemblages, except the presence of ortho-pyroxene in charnockite patches. The mineral chemistry studies revealed that mineral phases common in both rocks are chemically identical. Based on whole rocks geochemistry it is evident that both charnockite patches and host leptynites are iso-chemical. The adamellite composition, per-aluminous nature with positive Europium anomaly of host leptynite and charnockite patches suggests their co-genetic relation. Higher values of (La/Yb)N and (Gd/Lu)N of leptynites indicate the highly fractionated HREE pattern of leptynites. High concentrations of HREE in charnockite patches are due to the influence of fluid phase in granulite facies metamorphism. The charnockite patches are characterized by higher K2O content, higher HREE concentration, and less abundance of biotite indicating the in-situ growth of patchy charnockites.

The Paleoproterozoic thermal evolution of the north-Indian continental margin is debatable as the Lesser Himalaya granites are interpreted either as a subduction-zone volcanic arc or rift-related magmatism during Columbia assembly or disintegration processes, respectively. Integrated mineralogical, geochemical investigations, temperature estimates from Ti solubility in biotite and zircon, and computational phase equilibria modeling from the Wangtu Gneissic complex (WGC), Himachal Himalaya reveal peraluminous nature for most of the WGC rocks that crystallize at ~650 °C and ~ 1.0–1.1 GPa. The U-Pb ages from magmatic zircons from the WGC exhibit two prominent age clusters at 1867 Ma and 2487 Ma. The U-Pb zircon data and modeled phase equilibria for metasedimentary rock reveal the generation of S-type peraluminous magma parental to the WGC, through melting of preexisting supracrustal rocks at ~1800 Ma, within ~850–900 °C and 1.1–1.2 GPa, similar to P-T conditions observed in modern-day subduction zone settings. The results obtained from this study reveal the north Indian continental margin was an active subduction zone during the Paleoproterozoic Columbia supercontinent assembly that extends across the Himalayas.

As the relation between garnet chemistry and metamorphic condition is traditional, the importance of garnet in metamorphic petrology needs to be better understood. Garnet group of minerals provide information in terms of chemical composition, by utilizing which one can estimate the metamorphic P-T conditions, under which a rock has been formed. Khondalites, leptynites, charnockites and patchy charnockites are the important rocks exposed around Digapahandi area of Eastern Ghat Mobile Belt (EGMB) and garnet is a common mineral phase present in all these rock type. Hence garnets from each rock type were selected (total 39 points) for electron probe micro analysis (EPMA). A distinct compositional variation exists amongst garnets from different rocks, with abundant almandine content (67.845–88.028) and subordinate amount of pyrope, spessartine and grossular. These compositional variations in garnets from the different rocks of EGMB are not only due to variation in bulk rock composition but also due to distinct temperature and pressure conditions of metamorphism of these rocks. There is also a discrete variation in garnet composition from core to rim in each rock type. Growth zoning have developed in garnets from charnockites and khondalites while diffusion zoning is observed in the garnets from leptynites and patchy charnockites. Garnets in charnockites and khondalites are believed to have formed by prograde metamorphism during peak metamorphism. However garnets in leptynites and patchy charnockites are opined to have originated by retrograde metamorphism during exhumation following peak metamorphism.

Eastern Ghats Belt is a poly-metamorphic terrain, represents high to ultra-high temperature metamorphic condition. The temperature dominant metamorphism has proved by the presence of spinel bearing mineral assemblages like, spinel + quartz, spinel + sillimanite + cordierite, sapphirine + spinel etc. Present paper focus on spinel mineral chemistry, which present as a common phase in both leptynite and khondalite rocks around Digapahandi area of Eastern Ghats Belt. Petrographic observation confirms the presence of two different types of spinel in the rock. Coarse grain anhedral spinels are common in khondalites whereas in leptynites both euhedral and anhedral small spinels are present. EPMA analysis indicates that there is a distinct compositional variation in the spinel from both the rocks, although all the spinel falls under the category of Alrich spinel series. These evidences clearly reflect that both the rocks have suffered diverse metamorphic history.