Boulder weathering in McMurdo Dry Valleys, Antarctica

Abstract

Earth's dynamic surface undergoes a continuous cycle of mountain building and denudation. One of the important links in this cycle is the break-up and comminution of the rocks that allows for effective transportation of debris by surface processes. The starting and end points in this transformation are well known: bedrock and boulders on one end and silt and clay on the other. However, the existing knowledge of the rates and processes responsible of the intermediate steps is currently limited. To fill this gap in knowledge we studied boulders and their weathering products in the McMurdo Dry Valleys, Antarctica, that have been weathering for hundreds of thousands of years sub-aerially exposed at the ground surface. Our study boulders of locally distinct lithology have trails of rock fragments leading downhill revealing the rate of weathering and subsequent transport rate of the fragments. The rock fragments emanate from the source boulder and decrease in size as the distance downslope increases. We measured the fragment sizes and distances for various lithologies on varying slope angles. We found that large fragments up to 0.4m in diameter can be transported up to 60m downslope by unknown processes. The total length of the fragment trail increases with the slope angle. The maximum transport distances of sandstone boulders are approximately 10 times longer than other lithologies, which may be explained by the larger observed fragment sizes of the sandstones. On the other hand measurements of the smaller, generally less than 0.04m diameter fragments that are transported by wind, revealed much shorter transport distances (<10m). To gain insights of the boulder and resulting fragment weathering rates we constructed a boulder weathering-fragment transport computer model. The model is based on simple rules and probabilities that describe the weathering and transportation. The model is constrained by the observed fragment size distribution, fragment distribution in space, fragment size/distance relationship, and other observed properties of the fragments. This model allows us to determine the weathering rates of the source boulder and the resulting fragments on the ground, and the fragment size dependent transport rate that are consistent with the observations. Our modeling suggests that the boulders on average spall a fragment over 250 times more often than the resulting fragments re-break and that the currently observed fragment transport rate is consistent with our modeled long term average transport rate.