Air temperatures (Ta) are rising in a changing climate, increasing extreme temperature events. Examining how Ta increases are influencing extreme temperatures at the soil surface and belowground in the soil profile can refine our understanding of the ecological consequences of rising temperatures. In this paper, we validate surface and soil temperature (Ts: 0–100-cm depth) simulations in the SOILWAT2 model for 29 locations comprising 5 ecosystem types in the central and western USA. We determine the temperature characteristics of these locations from 1980 to 2015, and explore simulations of Ta and Ts change over 2030–2065 and 2065–2100 time periods using General Circulation Model (GCM) projections and the RCP 8.5 emissions scenario. We define temperature extremes using a nonstationary peak over threshold method, quantified from standard deviations above the mean (0-σ: an event >∼\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$>\sim $\end{document} 51% of extreme events; 2-σ:>∼98%\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$\sigma :>\sim 98\%$\end{document}). Our primary objective is to contrast the magnitude (∘C) and frequency of occurrence of extreme temperature events between the twentieth and twenty-first century. We project that temperatures will increase substantially in the twenty-first century. Extreme Ta events will experience the largest increases by magnitude, and extreme Ts events will experience the largest increases by proportion. On average, 2-σ extreme Ts events will increase by 3.4 ∘C in 2030–2065 and by 5.3 ∘C in 2065–2100. Increases in extreme Ts events will often exceed + 10 ∘C at 0–20 cm by 2065–2100, and at 0–100 cm will often exceed 5.0 standard deviations above 1980–2015 values. 2-σ extreme Ts events will increase from 0.9 events per decade in 1980–2015 to 23 events in 2030–2065 and 38 events in 2065–2100. By 2065–2100, the majority of months will experience extreme events that co-occur at 0–100 cm, which did not occur in 1980–2015. These projections illustrate the non-analog temperature increases that ecosystems will experience in the twenty-first century as a result of climate change.

中文翻译：

---

气候变化导致的空气、地表和地下温度极端事件的非模拟增加

气温 (Ta) 在不断变化的气候中上升，增加了极端温度事件。研究 Ta 的增加如何影响土壤剖面中土壤表面和地下的极端温度，可以加深我们对温度升高的生态后果的理解。在本文中，我们验证了 SOILWAT2 模型中的表面和土壤温度（Ts：0-100 厘米深度）模拟，适用于美国中部和西部的 29 个地点，包括 5 个生态系统类型。我们确定了这些位置从 1980 年到 2015 年的温度特征，并使用一般循环模型 (GCM) 预测和 RCP 8.5 排放情景探索了 2030-2065 和 2065-2100 时间段内 Ta 和 Ts 变化的模拟。我们使用非平稳峰值阈值方法定义极端温度，从高于平均值的标准偏差量化 (0-σ: an event >∼\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \ usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$>\sim $\end{document} 51% 的极端事件；2-σ:>∼98%\documentclass [12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt } \begin{document}$\sigma :>\sim 98\%$\end{document})。我们的主要目标是对比 20 世纪和 21 世纪极端温度事件的量级 (∘C) 和发生频率。我们预计气温将在 21 世纪大幅上升。极端 Ta 事件将经历最大的震级增长，极端 Ts 事件将经历最大比例的增长。平均而言，2-σ 极端 Ts 事件将在 2030-2065 年增加 3.4 ∘C，在 2065-2100 年增加 5.3 ∘C。到 2065-2100 年，极端 Ts 事件的增加通常会在 0-20 厘米处超过 + 10 ∘C，而在 0-100 厘米处，通常会超过 1980-2015 年值的 5.0 标准偏差。2-σ 极端 Ts 事件将从 1980-2015 年的每十年 0.9 个事件增加到 2030-2065 年的 23 个事件和 2065-2100 年的 38 个事件。到 2065-2100 年，大部分月份将经历在 0-100 厘米处同时发生的极端事件，这在 1980-2015 年没有发生。