Landscape Response to Rapid Uplift in Southwestern Taiwan: Insights from Denudation Rates Measurement from Cosmogenic 10Be (Meteoric)/9Be Ratios and Morphometric Indices

Located at an active arc-continent collision zone subject to a tropical climate, the Taiwan mountain belt is characterized by intense tectonic activity, resulting in rapid landscape evolution. In the Western Foothills of southwestern Taiwan, geodetic data reveal rapid surface deformation during periods of low seismicity, where the upper crust is dominated by mudstone lithology. However, mapped active structures do not fully explain the observed sharp deformation gradients and uplift patterns. This discrepancy motivates an evaluation of how strain is accommodated across different timescales and whether the present-day deformation reflects persistent long-term kinematics or transient processes.

Using the ratio of meteoric ¹⁰Be to mineral-weathered ⁹Be measured from five river-sediment samples collected from watersheds with distinct short-term uplift rates, spatial variations in basin-scale denudation rates (D_met) and their relationship to short-term uplift are evaluated. Although meteoric-¹⁰Be-derived denudation rates are particularly suitable for quartz-poor regions such as southwest Taiwan, the method relies on several assumptions that require validation. To assess its applicability, additional samples were collected from watersheds in the Central Range east of the study area, where in situ ¹⁰Be-derived denudation rates (D_insitu) are available.

In the Western Foothills area, D_met successfully captures large-basin denudation (0.77 ± 0.07 mm/yr) as the integrated signal of sub-basin denudation rates (average of 0.74 ± 0.01 mm/yr). Across two regions, D_met values are systematically lower in the Western Foothills than in the Central Range (5.8 - 7.4 mm/yr, with an outlier of 32 mm/yr), reflecting contrasts in lithology, climate setting, and topographic relief. In the Central Range, D_insituvalues (0.2-4.5 mm/yr) differ from D_met, suggesting that potential grain-size differences between the two methods lead to distinct sediment transport behaviors. Nevertheless, D_met remains informative by reproducing basin–sub-basin integration in the Western Foothills and distinguishing denudation regimes between regions.

Within the Western foothills, D_met correlates weakly with uplift rate, slope, and relief. The normalized channel steepness index (k_sn­) shows an unexpectedly weak to negative relationship with D_met. This pattern suggests that meteoric ¹⁰Be-derived denudation rates might not represent short-term surface deformation rates or are integrated over timescales that differ from those represented by geomorphic indices. This likely reflects transient surface adjustments rather than steady-state conditions. In contrast, D_met seems to positively correlate with the extent of barren-land (badland) surfaces developed in weak mudstone formation, suggesting first-order control on basin-averaged meteoric ¹⁰Be inventories. Although badlands have been proposed to be associated with rapid erosion in this region, the correspondence between their development timescale and the integration timescale of meteoric ¹⁰Be derived denudation remains uncertain.

Future work will expand sampling across additional basins spanning a wider range of badland extent and uplift signatures to test the robustness of these relationships and refine the link between short-term deformation and longer-term surface response. Additional analyses will quantify meteoric ¹⁰Be inventories on barren land and vegetated hillslopes to evaluate differences in meteoric ¹⁰Be retention across contrasting hillslope environments, thereby refining the applicability and sensitivity of the methods to hillslope transport processes.