山区夏季地表温度的影响因素——以泰山为例

Abstract

PDF HTML阅读 XML下载 导出引用 引用提醒 山区夏季地表温度的影响因素——以泰山为例 DOI: 10.5846/stxb201310312626 作者: 作者单位: 南京大学国际地球系统科学研究所,南京大学国际地球系统科学研究所,南京大学建筑与城市规划学院,南京大学国际地球系统科学研究所,南京大学国际地球系统科学研究所,南京大学建筑与城市规划学院 作者简介: 通讯作者: 中图分类号: 基金项目: 国家自然科学基金资助项目（31170444）；中央高校基本科研业务费专项资助 Analysis of factors affecting mountainous land surface temperature in the summer：a case study over Mount Tai Author: Affiliation: Nanjing University,International Institute for Earth System Sciences,,,, Fund Project: 摘要 | 图/表 | 访问统计 | 参考文献 | 相似文献 | 引证文献 | 资源附件 | 文章评论 摘要:以泰山为例，应用夏季的Landsat 5的TM6为基本数据源，基于单窗算法定量反演了泰山地表面温度（LST），在此基础上首先探讨了LST与地形因子的关系，然后比较了归一化水汽指数（NDMI）和归一化植被指数（NDVI）在表达山区LST上的效力，最后利用逐步回归分析法，构建出LST与地形因子、NDMI的回归方程，应用偏相关系数，得出各个因子对LST的影响程度。结果表明：1）在地形因子中，影响LST的主要因素是海拔，随海拔升高呈自然对数形式降低，相比而言，坡度、坡向以及太阳入射能量的影响则很小；2）在没有水体时，NDVI与NDMI都能有效地表达山区的LST，LST与NDVI间是二次项负相关关系，与NDMI间是线性负相关关系，在表达LST上NDMI比NDVI更有效；3）综合分析表明，地表水汽特征是其表面温度最主要的影响因素，其次是海拔。研究结果将为山区地表温度空间分异性特征及形成机制的研究提供科学的参考。 Abstract:Most of the previous studies related to land surface temperature (LST) are mainly focused on the investigation of urban heat island; however, little has been done on the mountainous area that are usually far away from cities. In this study, using the Landsat 5 Thematic Mapper (TM) at Mount Tai, firstly, the LST was retrieved based on the Mono-window Algorithm; and then the impacts on the LST from several factors including the topography, normalized difference moisture index (NDMI) and normalized difference vegetation index (NDVI) were analyzed though correlation analysis; accordingly, the regression equation between LST and topographic factors as well as NDMI was developed by stepwise regression analysis, the variable coefficients in the regression equation were interpreted using nonstandard regression coefficient, and then the impact of each factor on LST was quantized by partial correlation coefficient. The results show that: 1) In summer, elevation significantly affects the LST and has a significantly negative natural logarithm correlation rather than a negative linear correlation with it. However, the influence of slope, aspect and incident solar energy is not very significant, LST has a weak correlation with each of them; 2) NDVI and NDMI effectively express LST in mountainous areas if water is absent on surface. LST and NDVI have a negative quadratic correlation. In addition, with the increase of NDVI, the LSTs over areas covered by dense vegetation (NDVI > 0.5) will appear a "saturation" phenomenon. Meanwhile, LST and NDMI have a simple but stable negative linear correlation. When compared with NDVI, NDMI is more effective and more applicable at a large scale for the expression of LST;3) The comprehensive analysis shows that land surface moisture characteristic is the main factor affecting the LST, and then followed by the elevation. In comparison with the impacts of these two primary factors, those from the other factors are relatively insignificant. These results will provide important information on the examination of the spatial pattern and mechanism of LST across mountainous areas. 参考文献 相似文献 引证文献