Abstract Understanding the changes in Greenland's temperature is important for assessing and predicting the mass of the Greenland ice sheet, which plays an important role in sea level rise. In this study, we analyzed the annual and seasonal coastal Greenland's temperatures during the period 1952–2017 (focusing on the period 2013–2017) based on a dataset obtained from the Danish Meteorological Institute (DMI). Overall, the annual coastal Greenland's temperature increased during 1952–2017 at a rate of 0.23 °C decade−1, especially in the southeastern (0.70 °C decade−1) and northern (0.42 °C decade−1) regions of the island. From the changes in the seasonal coastal Greenland's composite temperature (CT), winter exhibited the largest change rate (0.28 °C decade−1), and the summer CT increased by 0.25 °C decade−1, while the spring CT increased by 0.17 °C decade−1 with less variation. The temperature increase accelerated during 2013–2017 according to Mann-Kendall (M-K) tests, especially in the northeastern and northern regions of the island. The seasonal temperature change of the whole island decreased in the following order: annual > autumn > summer > winter > spring. We also analyzed the annual inland temperature change during the period 1997–2017 based on a dataset obtained from the Greenland Climate Network; the results indicated that the inland temperature increased by 0.13 °C decade−1. Pearson correlation analysis was used to determine the teleconnection relationship between the coastal temperatures and large-scale atmosphere-ocean climate indexes, and we found that the Greenland Blocking Index (GBI), Atlantic Multidecadal Oscillation (AMO), Tropical Northern Atlantic Index (TNA), North Tropical Atlantic Index (NTA), Caribbean Index (CAR), Atlantic Meridional Mode (AMM), East Atlantic (EA) and Western Hemisphere warm pool (WHWP) have significant positive correlations with the coastal temperature in most months, except in February and May. However, the North Atlantic Oscillation (NAO), Arctic Oscillation (AO) and Eastern Asia/Western Russia (EAWR) show significant negative correlations with temperature. Overall, there exists a time lag effect between the climate indexes (except for the GBI, AO and NAO) and temperature. From the application of the random forest model, we found that the GBI, NAO, CO2, AMO, N2O, SF6, CH4, and Northern Oscillation Index (NOI) are the most important variables that influenced the CT changes during 1979–2017. Finally, we calculated the contribution rates of the most important variables to temperature change during the period 1979–2017 and showed that the contribution rates of the GBI, CO2 and NOI to temperature change were 47.30%, 35.68%, and 17.02%, respectively.

中文翻译：

---

格陵兰岛近五年升温速度加快

摘要 了解格陵兰岛的温度变化对于评估和预测格陵兰冰盖的质量具有重要意义，而格陵兰冰盖在海平面上升中起重要作用。在这项研究中，我们根据从丹麦气象研究所 (DMI) 获得的数据集分析了 1952 年至 2017 年期间（重点是 2013 年至 2017 年期间）格陵兰岛的年度和季节性沿海温度。总体而言，1952-2017 年间格陵兰沿海地区的年温度以 0.23 °C 十年-1 的速率上升，特别是在该岛的东南部（0.70 °C 十年-1）和北部（0.42 °C 十年-1）地区。从季节性沿海格陵兰岛的复合温度（CT）变化来看，冬季变化率最大（0.28°C十年-1），夏季CT增加0.25°C十年-1，而弹簧 CT 增加了 0.17 °C 十进制−1，变化较小。根据曼肯德尔 (MK) 测试，2013 年至 2017 年期间温度升高加速，特别是在该岛的东北部和北部地区。全岛季节气温变化递减顺序为：年>秋季>夏季>冬季>春季。我们还基于从格陵兰气候网络获得的数据集分析了 1997-2017 年期间内陆温度的年度变化；结果表明，内陆温度增加了 0.13 °C 十年-1。利用皮尔逊相关分析确定沿海温度与大尺度大气-海洋气候指数之间的遥相关关系，我们发现格陵兰阻塞指数（GBI）、大西洋年代际振荡 (AMO)、热带北大西洋指数 (TNA)、北热带大西洋指数 (NTA)、加勒比指数 (CAR)、大西洋经向模式 (AMM)、东大西洋 (EA) 和西半球暖池 (WHWP) 有除 2 月和 5 月外，大多数月份与沿海温度呈显着正相关。然而，北大西洋涛动（NAO）、北极涛动（AO）和东亚/俄罗斯西部（EAWR）与温度呈显着负相关。总体而言，气候指数（GBI、AO和NAO除外）与温度之间存在时滞效应。从随机森林模型的应用中，我们发现GBI、NAO、CO2、AMO、N2O、SF6、CH4和北方涛动指数（NOI）是影响1979-2017年CT变化的最重要变量。最后，