Abstract Urban trees play a key role in alleviating elevated summertime land surface temperatures in cities. However, urban landscape influences the capacity of urban trees to mitigate higher temperatures. We propose that both developed land characteristics and tree cover should be considered to accurately estimate the mitigation effects of canopy cover. We subclassified original land cover based on the canopy cover ratio to capture the within-land cover heterogeneity. We selected two coastal cities with different summertime climatic conditions: Seattle, Washington, USA, and Baltimore, Maryland, USA. We used Landsat-based grid cells (30 m × 30 m) as our spatial analytical unit, with corresponding land surface temperature, canopy area, canopy compactness, population size, and National Land Cover Database (NLCD)-based land cover group. We first used grouped boxplots, Kruskal–Wallis H tests, and post-hoc multiple comparison tests to detect the distribution of land surface temperatures by the land cover group. We then introduced statistical models to test the group effects on the relationship between land surface temperatures and canopy cover variables. We found: (1) land surface temperature increases with level of development, (2) land surface temperature decreases with canopy cover level, (3) the magnitude of the mitigation effects from canopy area differs based on development level and current canopy cover, (4) the differing efficacies of canopy area in decreasing land surface temperature follows a nonlinear threshold relationship, and (5) compactness of canopy cover was not significant in reducing the land surface temperature. These findings suggest the importance of considering heterogeneous canopy cover within developed land cover classes in urban heat island research. Tree planting strategies need to consider the nonlinear relationships between tree canopy cover and land surface temperature alongside environmental equity concerns.

中文翻译：

---

城市景观异质性影响树冠与地表温度的关系

摘要 城市树木在缓解城市夏季地表温度升高方面发挥着关键作用。然而，城市景观会影响城市树木缓解高温的能力。我们建议应同时考虑已开发的土地特征和树木覆盖，以准确估计冠层覆盖的缓解效果。我们根据冠层覆盖率对原始土地覆盖进行了细分，以捕获土地内部覆盖的异质性。我们选择了两个夏季气候条件不同的沿海城市：美国华盛顿州西雅图和美国马里兰州巴尔的摩。我们使用基于 Landsat 的网格单元 (30 m × 30 m) 作为我们的空间分析单元，具有相应的地表温度、冠层面积、冠层紧密度、人口规模和基于国家土地覆盖数据库 (NLCD) 的土地覆盖组。我们首先使用分组箱线图、Kruskal-Wallis H 检验和事后多重比较检验来检测土地覆盖组的地表温度分布。然后我们引入了统计模型来测试组对地表温度和冠层覆盖变量之间关系的影响。我们发现：(1) 地表温度随开发水平升高，(2) 地表温度随冠层覆盖水平降低，(3) 冠层面积减缓效应的大小因开发水平和当前冠层盖度而异，( 4）不同的冠层面积降低地表温度的效应遵循非线性阈值关系，（5）冠层覆盖的紧密度对降低地表温度不显着。这些发现表明在城市热岛研究中考虑已开发土地覆盖类别内的异质冠层覆盖的重要性。植树策略需要考虑树冠覆盖和地表温度之间的非线性关系以及环境公平问题。