Evolution of sandstone peak-forest landscapes - insights from quantifying erosional processes with cosmogenic nuclides

Abstract

The sandstone peak-forest landscape in Zhangjiajie UNESCO Global Geopark of Hunan Province, China, is characterized by >3000 vertical pillars and peak walls of up to 350 m height, representing a spectacular example of sandstone landform variety. Few studies have addressed the mechanisms and timescales of the longer-term evolution of this landscape, and have focused on fluvial incision. We use in situ cosmogenic nuclides combined with GIS analysis to investigate the erosional processes contributing to the formation of pillars and peak-forests, and discuss their relative roles in the formation and decay of the landscape. Model maximum-limiting bedrock erosion rates are the highest along the narrow fluvial channels and valleys at the base of the sandstone pillars (~83–122 mm kyr−1), and lowest on the peak wall tops (~2.5 mm kyr−1). Erosion rates are highly variable and intermediate along vertical sandstone peak walls and pillars (~30 to 84 mm kyr−1). Catchment-wide denudation rates from river sediment vary between ~26 and 96 mm kyr−1 and are generally consistent with vertical wall retreat rates. This highlights the importance of wall retreat for overall erosion in the sandstone peak-forest. In combination with GIS-derived erosional volumes, our results suggest that the peak-forest formation in Zhangjiajie commenced in the Pliocene, and that the general evolution of the landscape followed our sequential refined model: (i) slow lowering rates following initial uplift; (ii) fast plateau dissection by headward knickpoint propagation along joints and faults followed by; (iii) increasing contribution of wall retreat in the well-developed pillars and peak-forests and a gradual decrease in overall denudation rates, leading to; (iv) the final consumption of pillars and peak-forests. Our study provides an approach for quantifying the complex interplay between multiple geomorphic processes as required to assess the evolutionary pathways of other sandstone peak-forest landscapes across the globe.