Fill terrace formation and preservation histories along the Middle Kali Gandaki from digital and field investigations

Current debate exists on whether the depositional processes forming valley fill deposits in the Lesser Himalaya can be linked across multiple Himalayan river systems. Some studies have attributed the deposition of 100s of metres of sediment fill in the Lesser Himalaya to short-lived, catastrophic filling events. These events are thought to originate from large rock-slope failures and glacial lake outburst floods. Other studies have suggested that the deposition of fill on such large scales is a long-term process caused by fluctuations in the Indian Summer Monsoon (ISM), resulting in the oversupply of alluvial sediment into a transport-limited system. Cross-catchment patterns of terrace formation that might help to distinguish between these end-member driving mechanisms are limited by an incomplete record of fill terraces in the Lesser Himalaya, with multiple reaches having not yet seen detailed sedimentological study. Here, we address this gap by investigating a flight of fill terraces along the Middle Kali Gandaki River upstream of the town of Kushma, central Nepal. Terraces along this reach are preserved up to 400 m above the modern river channel. We build on previous studies of these terraces by identifying the stratigraphy of the terrace fill deposits and by recording clast lithology and morphology of the sediment preserved beneath each terrace level. We find the terraces to be predominantly composed of a coarse debris fill, with little to no visible stratigraphy, apart from occasional inverse grading from gravel to coarse boulder conglomerate. We identify five terrace levels and find differences in clast morphologies between T1, T2, T3, and T5, and a difference in clast lithology between T2 and T3. We therefore suggest that there have been at least four major filling and re-incision events along this reach, with the possibility of a fifth undocumented filling event forming T4. We also recreate approximate palaeo-valley floors for each terrace level to measure palaeo-valley widths. We interpret that multiple periods of extensive sediment aggradation and incision which led to the formation of five major terrace levels along the Middle Kali Gandaki River were driven by intensified monsoons leading to significant periods of sediment oversupply, contrasting the interpretations of catastrophic debris fill made along other rivers in the catchment. We attribute a decrease in valley width over time to high denudation rates near the MCT. We investigate terrace preservation along the study reach by calculating the percentage of preserved terrace area compared to the approximate area of the abandoned valley floor for each terrace level. We find that terrace levels T3 and T4 are poorly preserved compared to T1 and T2. We suggest post-depositional cementation to be a primary control, with indurated older terrace deposits acting as bedrock, rapidly reducing valley width so that the river can erode the entire floodplain more easily and thus limiting terrace preservation.