The land surface temperature (LST) of most cities is rising steadily due to both human activities and global climate change, affecting the thermal comfort of cities and threatening the physical health of their inhabitants. Investigating surface temperature features and their affecting elements is therefore essential to improve the thermal environment of cities. Due to differences in surface temperature characteristics in different climatic belts. However, there is still a lack of research on how the local climate zones (LCZs) of different climate belts are affected by natural and social causes. According to the Koppen-Geiger climate classification system, we selected three representative cities from each of the four macroclimatic belts and adopted multiple linear stepwise regression and boosted regression trees (BRT) to systematically explore the linear relationships, relative impacts and marginal effects of normalized difference vegetation index (NDVI), modified normalized difference water index (MNDWI), population density, and road nuclear density within urban LCZs on LST. The results indicate that (1) LCZs with significant differences in LST among the four global climatic belts account for more than 95 % (P < 0.05), demonstrating that LCZs can effectively differentiate LST based on different land surface cover types. This study can delve into the relationships between the four influencing factors and LST based on LCZs. (2) Primary control factors regulating LST differ in different climate belts. The relative effects of NDVI and MNDWI on LST are greater in the arid and temperate belts, with each 0.1 increase in NDVI and MNDWI producing a cooling effect of more than 0.40 °C and 0.92 °C, respectively. Ventilation corridors created by increased road core density produced a cooling effect of 1 °C or more in the cold belt, with the most pronounced cooling effect. (3) When natural factors are higher and social factors are lower, LST cannot be minimized. It was found that LST was minimized when NDVI, MNDWI, population density and road nuclear density were controlled above 0.4, 0.1–0.2, 10,000–50,000 and 1500–2000 respectively. Our study will provide targeted measures for LCZ-based mitigation of urban thermal environments in various macroclimatic belts.

中文翻译：

---

全球典型城市局部气候区内地表温度的多因素影响


由于人类活动和全球气候变化，大多数城市的地表温度（LST）稳步上升，影响城市的热舒适度并威胁居民的身体健康。因此，研究地表温度特征及其影响因素对于改善城市热环境至关重要。由于不同气候带地表温度特征存在差异。然而，对于不同气候带的局地气候带（LCZ）如何受到自然和社会原因的影响仍缺乏研究。根据Koppen-Geiger气候分类系统，我们从四个大气候带中各选取了三个代表性城市，采用多元线性逐步回归和提升回归树（BRT）系统地探讨了归一化差异的线性关系、相对影响和边际效应LST 上城市 LCZ 内的植被指数 (NDVI)、修正的归一化水分指数 (MNDWI)、人口密度和道路核密度。结果表明：（1）全球四大气候带中LST差异显着的LCZ占95%以上（P<<0.05），表明LCZ能够根据不同的地表覆盖类型有效区分LST。本研究可以基于LCZ深入探讨四个影响因素与LST之间的关系。 (2)不同气候带地表温度的主控因素不同。 NDVI和MNDWI对地表温度的相对影响在干旱和温带地区更大，NDVI和MNDWI每增加0.1，分别产生超过0.40℃和0.92℃的降温效果。 增加道路核心密度形成的通风廊道在冷带中产生1℃以上的降温效果，且降温效果最为明显。 (3)当自然因素较高、社会因素较低时，LST不能最小化。研究发现，当NDVI、MNDWI、人口密度和道路核密度分别控制在0.4、0.1~0.2、10,000~50,000和1500~2000以上时，LST最小。我们的研究将为基于 LCZ 的各种大气候带城市热环境缓解提供有针对性的措施。