We show that climate and topography control the spatial distribution of stable isotope values on the South Island of New Zealand, based on a spatially dense (n = 193) river isotopic survey. Our data show a δ¹⁸O minimum in isotope values east of the Southern Alps that demonstrates topographically driven continentality associated with the Southern Alps, which intersect the prevailing, moisture-laden westerlies. Our data define a South Island surface water line of δ²H = 8.17 (±0.26) × δ¹⁸O + 10.57 (±2.04), which is identical within 95% confidence intervals to the global and New Zealand meteoric water lines established from monthly precipitation samples. The observed river δ¹⁸O values are strongly correlated with annual temperature range and winter temperature. Strongest correlations are between δ¹⁸O and mean minimum winter temperatures (r > 0.7 for June, July, August), with gradients of 0.58–0.66‰ /°C. Based on a multiple regression analysis of δ¹⁸O against climate data, we present a river δ¹⁸O model and isoscape that demonstrate the control of continentality and moisture source on New Zealand surface water isotope spatial patterns. Model validation against previously published river samples shows skill in predicting river δ¹⁸O values (root-mean-square error = 0.83), confirming that the spatial variations in river δ¹⁸O (and δ²H) are robust to sampling period and reflect continental, precipitation source and temperature effects. Our data suggest that oxygen or hydrogen isotope paleoclimate proxies derived from rivers or open-system lakes on the South Island should be sensitive to winter temperature.

中文翻译：

---

新西兰南岛河流稳定同位素值的气候和地形控制

我们显示，根据空间密集的（n  = 193）河流同位素调查，气候和地形控制了新西兰南岛上稳定同位素值的空间分布。我们的数据显示δ ¹⁸ Ø最小同位素值的南阿尔卑斯山，演示与南阿尔卑斯山，相交盛行，充满水分的西风相关地形推动大陆度以东。我们的数据定义δ的南岛表面水管线² H = 8.17（±0.26）×δ ¹⁸ O + 10.57（±2.04），其内的95％置信区间是相同的，从每月建立的全球和新西兰大气水线降水样品。观察到的河水δ ¹⁸O值与年温度范围和冬季温度密切相关。最强相关性δ之间¹⁸ O和平均最低冬季温度（[R  > 0.7六月，七月，八月），具有0.58-0.66‰/℃的梯度。根据δ的多元回归分析¹⁸对抗气候数据O，我们提出了一个河流δ ¹⁸ O模型和isoscape演示新西兰地表水同位素空间模式大陆度和水分源的控制。在预测河流δ针对先前公布的河流样品显示技能模型验证^18个O值（根均方误差= 0.83），这证实了在河流δ的空间变化¹⁸ O（和δ ²H）对采样周期具有鲁棒性，并反映了大陆，降水源和温度的影响。我们的数据表明，来自南岛的河流或开放系统湖泊的氧或氢同位素古气候代理应该对冬季温度敏感。