中小尺度下植被冠层对屋顶表面温度的调控效应分析
作者简介:杨 若(1993-),女,广东河源人,硕士生,研究方向为空间分析与数据挖掘。E-mail: joengjoek@gmail.com
收稿日期: 2018-10-25
要求修回日期: 2019-03-15
网络出版日期: 2019-07-25
基金资助
地表载荷作用下MTINSAR城市地面沉降监测及时空多尺度演化规律挖掘(41671417)
Effect of Vegetation Canopy on Rooftop Surface Temperature at City Block and Building Scale
Received date: 2018-10-25
Request revised date: 2019-03-15
Online published: 2019-07-25
Supported by
MTINSAR-based Urban Land Subsidence Monitoring and Spatial Temporal Multi-scale Trend Characteristic Mining with the Impact of Ground Load, No.41671417
Copyright
随着城市化进程的加快,城市热岛问题日益严重,对人类健康和城市可持续发展产生了巨大威胁。植被可有效遮蔽阳光直射,并通过蒸腾作用降低气温,是改善局部热环境的重要途径之一。开展植被对建筑物温度的调控效应的研究,对于理解城市热岛成因、缓解城市热环境恶化等方面都有重要意义。然而,当前研究往往是在遥感影像的基础上进行的,缺乏植被结构信息,同时,受制于有限的空间分辨率,研究大多在城市尺度下开展。在中小尺度上定量地研究植被冠层密度对建筑物温度的影响仍然具有一定挑战性。鉴于此,本文使用激光雷达(Light Detection and Ranging, LiDAR)获取的高分辨率冠层密度数据,在楼间尺度和街区尺度下开展圣罗莎市三维植被结构与单体建筑物表面温度之间定量关系的研究,分析不同尺度下植被冠层的降温特征及其在局部环境中的降温贡献。结果表明:植被对建筑物的降温作用与其周围的冠层密度有密切关系:冠层密度需达到17%才能起到有效的降温作用,其中在中小尺度上冠层密度分别高于30%和40%时,能最大限度发挥植被的温度调控功能;当冠层密度相同时,2个尺度下的温度变化显著不同:随着冠层密度的增加,街区尺度下的屋顶温度比楼间尺度下的屋顶温度平均下降了0.89 ℃;中小尺度下的屋顶温度变化不仅受到其周围植被结构的影响,还与整体热环境状况有关。本文的研究思路与结果有助于在有限的城区土地资源上合理规划绿地建设,构建可持续的人类宜居环境。
杨若 , 敖祖锐 , 张晶 , 余洁 . 中小尺度下植被冠层对屋顶表面温度的调控效应分析[J]. 地球信息科学学报, 2019 , 21(7) : 1097 -1108 . DOI: 10.12082/dqxxkx.2019.180547
With the acceleration of urbanization, urban heat island (UHI) effect has become an increasingly serious problem, which poses a great threat to public health and urban sustainability. Vegetation can lower the air temperature by reflecting direct sunlight and through the process of evapotranspiration, and hence plays a key role in improving local thermal environments. Investigating the effect of vegetation on regulating building temperature is very useful for understanding the principle of urban heat island and mitigating the deterioration of urban thermal environment. However, most previous studies are based on remote sensing imagery, which lacks three-dimensional information on vegetation structure. Additionally, these studies are mainly carried out at the urban scale due to the limitation of spatial resolution. Therefore, it remains challenging to quantitatively investigate the effects of vegetation canopy structure on building temperature at small and medium scales. In this paper, we quantitatively investigated the relationship between the LiDAR-derived 3D vegetation structure (canopy density, CD) and the rooftop surface temperature (RST) at the city-block (medium) and individual building (small) scale. We improved the Building Thermal Functional Area model (BTFA). Considering the spatial and quantity characteristics of buildings in Santa Rosa, the optimal sizes of the small and medium thermal function areas were estimated. Then the vegetation canopy density around the buildings at two scales were calculated. The cooling capacity of CD was analyzed by nonlinear fitting model and other statistical methods. Moreover, we used spatial autoregression model to analyze the contribution of CD to lower the rooftop temperature under the interaction of various factors. Results show that the cooling effect of vegetation on buildings is closely related to the canopy density around them: the minimum threshold of 17% is required to achieve effective cooling effect, while 30% and 40% are the optimal thresholds at medium and small scales, respectively. Additionally, changes of RST vary at different scales with the same canopy density. The decrease of RST at the medium scale is on average 0.89 ℃ lager than that at the small scale. The findings suggest that the planning of urban green space should be considered comprehensively in different scales. Moreover, the RST changes at small and medium scales are affected by not only the vegetation structure nearby the buildings but also the overall thermal environment. The methods and results of this paper are helpful to better plan green spaces on the limited urban land resources and build a more sustainable human livable environment.
Fig. 1 DEM map of Santa Rosa, California, USA图1 美国加州圣罗莎市地形示意 |
Fig. 2 Research flowchart of the methods图2 总体技术路线 |
Fig. 3 Result of the G-function method for distance between buildings图3 建筑点距离分布G函数曲线 |
Fig. 4 Building thermal functional area (BTFA) model图4 BTFA模型示意 |
Fig. 5 Spatial distribution of land surface temperature zones of Santa Rosa in October 2013图5 2013年10月圣罗莎市地表温度等级划分 |
Fig. 6 Land use map of Santa Rosa in October 2013 ( The result of supervised classification )图6 2013年10月圣罗莎市土地利用类型(监督分类结果) |
Fig. 7 Fractional vegetation cover map of Santa Rosa in October 2013图7 2013年10月圣罗莎市植被覆盖度图像 |
Fig. 8 Canopy density image of Santa Rosa in fall of 2013图8 2013年秋季圣罗莎市冠层密度图像 |
Fig. 9 Building frequencies for each canopy density quartile observed in different temperature zones of BTFA_R50图9 BTFA_R50的4个温度区中建筑物数量在各自冠层密度分组中的比例 |
Fig. 10 Building frequencies for each canopy density quartile observed in different temperature zones of BTFA_R300图10 BTFA_R300的4个温度区中建筑物数量在各自冠层密度分组中的比例 |
Fig. 11 Smoothed scatter plot and nonlinear curve fitting of CD & RST in BTFA_R50图11 BTFA_R50冠层密度与屋顶表面温度的高密度散点图与非线性拟合 |
Fig. 12 Smoothed scatter plot and nonlinear curve fitting of CD & RST in BTFA_R300图12 BTFA_R300冠层密度与屋顶表面温度的高密度散点图与非线性拟合 |
Tab. 1 Comparison of cooling amplitude between two scales in different canopy density intervals (℃)表1 2个尺度在不同冠层密度区间的降温程度比较 |
冠层密度区间 | 降温程度(BTFA_R50) | 降温程度(BTFA_R300) | 温差 |
---|---|---|---|
0~10% | -1.68 | -2.18 | 0.50 |
0~20% | -3.15 | -3.99 | 0.84 |
0~30% | -4.39 | -5.44 | 1.05 |
0~40% | -5.41 | -6.52 | 1.11 |
0~50% | -6.20 | -7.24 | 1.04 |
0~60% | -6.78 | -7.59 | 0.81 |
Tab. 2 Comparison of the statistics of OLS, SLM, and SEM models between two scales表2 BTFA_R50和BTFA_R300的OLS、SLM、SEM模型估计与检验结果对比 |
BTFA_R50 | BTFA_R300 | ||||||
---|---|---|---|---|---|---|---|
OLS | SLM | SEM | OLS | SLM | SEM | ||
Intercept | 31.85*** | 8.97*** | 30.60*** | 32.27*** | 8.48*** | 31.88*** | |
Canopy density | -12.52*** | -4.92*** | -6.12*** | -14.19*** | -4.34*** | -12.37*** | |
Spatial lag(ρ) | 0.73*** | 0.74*** | |||||
Spatial error(λ) | 0.83*** | 0.76*** | |||||
R2 | 0.57 | 0.87 | 0.88 | 0.56 | 0.84 | 0.85 | |
LogL | -8290.86 | -5492.41 | -5730.16 | -8341.49 | -6087.24 | -6132.52 | |
AIC | 16 585.7 | 10 993 | 11 468 | 16 687 | 12 182 | 12 273 |
注:***、**、* 分别表示0.001、0.01、0.05的显著性水平。 |
Tab. 3 Diagnostics for spatial dependence of the OLS model表3 OLS模型空间依赖性检验 |
统计量 | 统计值(BTFA_R50) | 统计值(BTFA_R300) |
---|---|---|
Moran's I (error) | 0.63*** | 0.68*** |
LM-lag | 4591.10*** | 4800.10*** |
LM-error | 4000.10*** | 4707.90*** |
Robust LM-lag | 823.70*** | 137.70*** |
Robust LM-error | 232.73*** | 45.55*** |
注:***、**、* 分别表示0.001、0.01、0.05的显著性水平。 |
[1] |
|
[2] |
|
[3] |
|
[4] |
|
[5] |
[
|
[6] |
[
|
[7] |
[
|
[8] |
[
|
[9] |
[
|
[10] |
|
[11] |
[
|
[12] |
|
[13] |
|
[14] |
|
[15] |
|
[16] |
|
[17] |
[
|
[18] |
|
[19] |
|
[20] |
|
[21] |
[
|
[22] |
|
[23] |
[
|
[24] |
[
|
[25] |
|
[26] |
|
[27] |
[
|
[28] |
[
|
/
〈 | 〉 |