Ground Surface Temperature (GST) is especially relevant in permafrost regions, such as the ice-free areas of the Antarctic Peninsula, to the understanding of environmental changes, where a long-term warming trend has been detected since 1950. To better understand GST regimes and the topoclimatic controlling factors, 20 iButtons were installed at sites according to elevation, exposure, curvature, and proximity to permanent snow, recording temperatures at 3-hour intervals from March 2019 to February 2020. Multiple Factorial Analysis (MFA) was used to evaluate the influence of these factors on GST parameters and to group the sensors based on their similarities. An analysis of daily temperatures was conducted to classify types of daily ground temperature regimes for use in a spatial model, developed using Discriminant Analysis (DA). As was predicted elevation was identified as the main controlling factor, with a negative correlation to the Mean Annual Ground Surface Temperature, ranging from 0.6 °C at 16 m a.s.l. to −2 °C at 254 m a.s.l., and a positive correlation with Freezing Degree Days, ranging from 438 (at 16 m a.s.l.) to 1042 (at 254 m a.s.l.). Snow cover duration is the second control factor highlighted, determining the duration of the freezing season, which was prolonged where snow cover persisted longer, resulting in a more pronounced insulating effect. The diversity of conditions was reinforced with the identification of seven types of daily GST regimes (three frozen, two unfrozen, and two with freeze–thaw), leading to the categorization of four annual distribution types. These were spatialized for Barton Peninsula using the DA model (90 % accuracy). The spatialization revealed a long frost season in proximity to snow patches, moderate frost season in areas above 160 m a.s.l., a short frost season with slow warming in areas ranging from 90 to 160 m a.s.l., and a short frost season with rapid warming in areas below 90 m a.s.l.

中文翻译：

---

地表温度状况受南极洲海洋无冰地区的地形和积雪控制

地表温度 (GST) 在永久冻土地区（例如南极半岛的无冰地区）尤其与了解环境变化相关，自 1950 年以来已检测到长期变暖趋势。为了更好地了解 GST 制度和地形气候控制因素，根据海拔、暴露度、曲率和与永久积雪的接近程度，在不同地点安装了 20 个 iButton，从 2019 年 3 月到 2020 年 2 月，每隔 3 小时记录一次温度。使用多重因子分析 (MFA) 进行评估这些因素对 GST 参数的影响，并根据传感器的相似性对传感器进行分组。对每日气温进行分析，对每日地面温度状况进行分类，以用于使用判别分析 (DA) 开发的空间模型。正如预测的那样，海拔被确定为主要控制因素，与年平均地表温度呈负相关，范围从 16 m asl 时的 0.6 °C 到 254 m asl 时的 -2 °C，与冰冻度呈正相关天，范围从 438 天（海拔 16 m）到 1042 天（海拔 254 m）。积雪持续时间是第二个强调的控制因素，决定了冰冻季节的持续时间，积雪持续时间越长，冰冻季节就越长，从而产生更明显的隔热效果。通过确定七种类型的每日商品及服务税制度（三种冻结、两种未冻结和两种冻融），加强了条件的多样性，从而对四种年度分配类型进行了分类。使用 DA 模型对巴顿半岛进行空间化（准确度为 90%）。空间化结果显示，雪地附近地区霜季较长，海拔160 m以上地区为中霜期，海拔90~160 m地区为短霜期，升温缓慢，海拔90~160 m地区为短霜期，升温较快。海拔低于 90 m