Exploring Controls on the Grain Size Distribution of Weathered Rock in the Luquillo Mountains, Puerto Rico with Landlab

Abstract:

The grain size distribution of sediment delivered to rivers depends on the conversion of rock to regolith through weathering. Understanding the controls on sediment grain size is essential because grain size may affect hillslope relief, fluvial processes, and channel steepness. This research aims to identify controls on sediment grain size in different parts of the Rio Blanco watershed, Puerto Rico. Most of the study area consists of quartz diorite, with minor areas of volcaniclastic rock. The watershed is in a transient state, with knickpoints at ~ 600 m dividing an upstream low relief landscape from a steeper landscape downstream. The erosion rate in the upper portion of the landscape is ~50 mm/kyr, and it doubles downstream of the knickpoint.

We illustrate results using the framework and geomorphic transport laws developed by Sklar et al. (2017) to model the grain-size distribution resulting from the transformation of rock to soil. The equations predict the grain-size distribution as a function of rock type, precipitation, temperature, and erosion rate. We tested the sensitivity of the degree of weathering to the different variables using the Python SALib library. The results show that erosion properties (magnitude of erosion rate, erosion rate thresholds to supply limited and kinetically limited weathering) are the most critical terms and have a larger control on weathering in this framework than lithology and climate parameters. We also tested the impact of differing erosion rates in the Rio Blanco on the grain size distribution. In areas of the watershed where erosion rates are higher, coarser grain sizes are predicted because the residence time of particles in soil is shorter. The difference in grain size between the faster and slower eroding portions of the watershed depends on the threshold erosion rates for weathering transitions. If steep portions of the landscape (>35 degrees) are modeled as having the minimum possible weathering given the local climate, a bimodal grain size distribution results. However, because only ~5% of the Rio Blanco watershed is steep, this does not result in a large input of coarse grains to the channel. Overall, given our best parameterization of the model, it does not predict large differences in the grain size distributions between the two portions of the watershed.