Over the past decades, luminescence has been widely used for dating sedimentary deposits. Several recent publications suggest luminescence signals can also be used to investigate fluvial transport. Here we explore what information luminescence signals yield in past and present sediment dynamics in the Rangitikei River (RR), New Zealand (Bonnet et al., 2019).
We present a dataset of 30 samples from fluvial terraces and modern river sediments of the RR. For each of the samples, we measured pIRIR luminescence signals of 300 individual sand-sized grains of feldspar (Reimann et al., 2012). We interpret results to evaluate differences between past and modern transport conditions, and to infer information on lateral input of bedrock particles in different river sections.
The information obtained from the single-grain analysis is incredibly rich, and requires new metrics for interpretation. To quantify the percentage of grains that were eroded from bedrock (or very old deposits) and re-deposited with minimal light-exposure, we identified grains for which the pIRIR signal is above 85% of full saturation (Wintle, 2006). For grains below this saturation threshold, we used the bootstrapped minimum age model (Galbraith et al.,1999; Cunningham and Wallinga, 2012) to determine the palaeodose, the best estimate of the natural radiation dose received by grains since their last deposition and burial event. For the modern deposits, we interpret the palaeodose to indicate the light-exposure of the best-bleached grains. Thereby, it provides a proxy of fluvial transport distance of the sediment grains.
For the modern river sediments we obtain palaeodoses between 2 and 6 Gy. A decreasing trend in palaeodose downstream suggests that part of the grains are transported through the entire system and are gradually bleached through light exposure during this process. The downstream trend in palaeodose of the RR is influenced by the connection of a major tributary, the Kawhatau River (KR), characterized by higher palaeodoses. Based on the observed trends, we estimate that the KR contributes three times more to modern sediment flux down the confluence than the upstream RR. Moreover, we observe that downstream of the confluence the percentage of saturated grains increase, which implies significant local input of bedrock particles from valley sides.
Data from recent (Holocene) autogenic fluvial terraces were acquired downstream the RR/KR confluence. They show a high to very high ratio of saturated grains (30-70%). We also document a downstream increasing trend of the percentage of saturated grains in these fluvial terraces, much stronger than for modern deposits. The maximum is observed for terraces at elevation of +28/+34 m, with an input of saturated grains that doubles over a distance of 100 km. As a consequence, saturated grains represent up to 70 % of the grain population in the most downstream sample. This implies a stronger lateral input of bedrock particles in the past, during recent incision of the river and a significant contribution of valley walls to the sediment flux of the RR, probably through landslides and/or lateral fluvial erosion.