Night-time minimum temperatures are increasing, faster than day-time maximum temperatures. They are also strongly connected to key outcomes such as health, vegetation, and crop productivity, and therefore changes in the distributions of night-time minimum temperatures have strong impacts on living systems. Night-time minimum temperatures are also related to teleconnections, which are indices that capture climate phenomenon occurring on a large spatial and temporal scale. As autocorrelated sequences, teleconnections are predictable and therefore inform distributions of weather in the near future. In this paper, we link four well-known teleconnections to night-time minimum temperatures in the American Southwest. These teleconnections are covariates for the mean process, but also for the variance, which is modelled as an exponential generalized autoregressive conditional heteroskedastic process (AR-EGARCH). The resulting models are used to estimate the probability of a low temperature event across the phases of each teleconnection, demonstrating the impact each teleconnection has on variability of night-time minimum temperatures.

夜间最低温度正在增加，快于白天最高温度。它们还与健康，植被和农作物生产力等关键成果紧密相关，因此夜间最低温度分布的变化会对生命系统产生强烈影响。夜间最低温度也与远程连接有关，远程连接是捕获在大时空尺度上发生的气候现象的指标。作为自相关序列，远程连接是可预测的，因此可以在不久的将来通知天气分布。在本文中，我们将四个著名的远程连接链接到美国西南部的夜间最低温度。这些遥相关是均值过程的协变量，也是方差的协变量，它被建模为指数广义自回归条件异方差过程（AR-EGARCH）。所得的模型用于估计每个远程连接各个阶段发生低温事件的概率，从而证明每个远程连接对夜间最低温度变化的影响。