This study reviews the Quaternary alluvial stratigraphy in three semi-arid river basins of western India i.e., lower Luni (Rajasthan), and Mahi and Sabarmati (Gujarat alluvial plains). On the basis of OSL chronologies, it is shown that the existing intra-valley lithostratigraphic correlations require a revision. The sand, gravel and mud facies are present during various times in the three basins, however, the fluvial response to climate change, and the resulting facies associations, was different in the Thar desert as compared to that at the desert margin; this makes purely lithostratigraphic correlations unviable. It is further shown that the rivers in the Thar desert were more sensitive to climate change and had small response times and geomorphic thresholds as compared to the desert-margin rivers. This is illustrated during the early OIS 1, when the Luni river in the Thar desert was dynamic and showed frequent variations in fluvial styles such as gravel bedload braided streams, sand-bed ephemeral streams and meandering streams, all followed by incision during the early Holocene. The coeval deposits in Sabarmati, however, only show a meandering, floodplain-dominated river.

Late Quaternary alluvial deposits in these basins unconformably overlie some older deposits that lack any absolute chronology. Based on the facies types and their associations, and the composition and architecture of the multistoried gravel sheets in the studied sections, it is suggested that older deposits are of pre-Quaternary age. This hypothesis implies the presence of a large hiatus incorporating much of the Quaternary period in the exposed sections

The calcretes in the Thar desert occur in a variety of settings, including the piedmonts, sheetwash aggraded plains; and this study adds calcretes in regolith and colluvio-alluvial plains to the group of settings in which calcretes occur in the region. Field logs, morphological details and analytical data such as petrographic, cathodoluminescence and geochemical characteristics are described along with a discussion on their implications. Sand dunes and sandy plains dating to < 20 ka have weakly developed calcretes. The better-developed calcrete horizons occur in piedmonts, interdunes or in areas that have sufficient groundwater. Deep sections in the region show phases of calcrete development in aeolian sand aggradation at ∼ 150, ∼ 100, ∼ 60 and 27–14 ka. The extensive sheetwash plains have mature calcretes and date to mid-Pleistocene. Our studies indicate that these calcretes represent a hybrid process, where carbonate enrichment of the originally calcareous host occurred due to periodically raised groundwaters, and its differentiation into nodules occurred under subaerial environment i.e., after recession of groundwater. Deep sections also show a stack of discrete calcretes that developed in individual aggradation episodes with hiatuses as indicated by ESR dating results. Nodules display a multiplicity of carbonate precipi tation events and internal reorganization of calcitic groundmass. The process is accompanied by degradation and transformation of unstable minerals, particularly clays and with a neosynthesis of palygorskite.

The ancient calcretes are dated from the beginning of the Quaternary to ∼ 600 ka and show more evolved morphologies marked by brecciation, dissolution, laminar growth on brecciated surfaces, pisolites and several generations of re-cementation. Mica/chlorite schists and such other rocks are particularly vulnerable to replacement by carbonate. In an extreme case, replacement of quartzose sandstone was observed also. The presence of stretches of alluvio-colluvial plains in an area presently devoid of drainage bespeaks of occasional high-energy fluvial regime, under a semi-arid climate. The mid-Pleistocene period saw a shift towards more arid climate and this facilitated sheetwash aggradation. Finally, during the late Pleistocene, aggradation of aeolian sands indicated a progressively drier climate. However, this does not find its reflection in stable isotope data. The amount of carbonate in the form of calcretes is substantial. The present studies indicate that aeolian dust or rainwater are minor contributors to the carbonate budget. A more important source was provided by the pre-existing calcretes in the sheetwash aggraded plains and detrital carbonate in the aeolian sediments. The original source of carbonate in the region, however, remains unresolved and will need further investigations. Electron spin resonance protocols for the dating of calcretes were developed as a part of this study and the results accorded well with geological reasoning