<p>The Southern Alps of New Zealand are the expression of the oblique convergence between the Pacific and Australian plates, which move at a relative velocity of nearly 40 mm/yr. This convergence is accommodated by the range-bounding Alpine Fault, with a strike-slip component of ~30-40 mm/yr, and a shortening component normal to the fault of ~8-10 mm/yr. While strike-slip rates seem to be fairly constant along the Alpine Fault, throw rates appear to vary considerably, and whether the locus of maximum exhumation is located near the fault, at the main drainage divide, or part-way between, is still debated. These uncertainties stem from very limited data characterizing vertical deformation rates along and across the Southern Alps. Thermochronology has constrained the Southern Alps exhumation history since the Miocene, but Quaternary exhumation is hard to resolve precisely due to the very high exhumation rates. Likewise, GPS surveys estimate a vertical uplift of ~5 mm/yr, but integrate only over ~10 yr timescales and are restricted to one transect across the range.</p><p>To obtain insights into the Quaternary distribution and rates of exhumation of the western Southern Alps, we use new <sup>10</sup>Be catchment-averaged erosion rates from 20 catchments along the western side of the range. Catchment-averaged erosion rates span an order of magnitude, between ~0.8 and >10 mm/yr, but we find that erosion rates of >10 mm/yr, a value often quoted in the literature as representative for the entire range, are very localized. Moreover, erosion rates decrease sharply north of the intersection with the Marlborough Fault System, suggesting substantial slip partitioning. These <sup>10</sup>Be catchment-averaged erosion rates integrate, on average, over the last ~300 yrs. Considering that the last earthquake on the Alpine Fault was in 1717, these rates are representative of inter-seismic erosion. Lake sedimentation rates and coseismic landslide modelling suggest that long-term (~10<sup>3</sup> yrs) erosion rates over a full seismic cycle could be ~40% greater than our inter-seismic erosion rates. If we assume steady state topography, such a scaling of our <sup>10</sup>Be erosion rate estimates can be used to estimate rock uplift rates in the Southern Alps. Finally, we find that erosion, and hence potentially exhumation, does not seem to be localized at a particular distance from the fault, as some tectonic and provenance studies have suggested. Instead, we find that superimposed on the primary tectonic control, there is an elevation/temperature control on erosion rates, which is probably transient and related to frost-cracking and glacial retreat.</p><p>Our results highlight the potential for <sup>10</sup>Be catchment-averaged erosion rates to provide insights into the magnitude and distribution of tectonic deformation rates, and the limitations that arise from transient erosion controls related to the seismic cycle and climate-modulated surface processes.</p><p>&#160;</p><p>&#160;</p>
An early Holocene glacial advance in the Macaulay River valley, central Southern Alps, New Zealand
