Land surface temperature shaped by urban fractions in megacity region

The influence of spatial scales on surface fluxes is an interesting but not fully investigated question. This paper presents an analysis on the influence of spatial scales on surface fluxes in the north Tibetan Plateau based on eddy covariance (EC) and large aperture scintillometer (LAS) data at site Nagqu/BJ, combined with the land surface temperature (LST) and normalized difference vegetation index (NDVI) of moderate-resolution imaging spectroradiometer (MODIS). The analysis shows that sensible heat fluxes calculated with LAS data (H_LAS) agree reasonably well with sensible heat fluxes calculated with EC data (H_EC) in the rain and dry seasons. The difference in their footprints due to the wind direction is an important reason for the differences in H_EC and H_LAS. The H_LAS are statistically more consistent with H_EC when their footprints overlap than when their footprints do not. A detailed analysis on H_EC and H_LAS changes with net radiation and wind direction in rain and dry season indicates that the spatial heterogeneity in net radiation created by clouds contributes greatly to the differences in H_EC and H_LAS in short-term variations. A significant relationship between the difference in footprint-weighted averages of LST and difference in H_EC and H_LAS suggests that the spatial heterogeneity in LST at two spatial scales is a reason for the differences in H_EC and H_LAS and that LST has a positive correlation with the differences in H_EC and H_LAS. A significant relationship between the footprint-weighted averages of NDVI and the ratio of sensible heat fluxes at two spatial scales to net radiation (H/Rn) in the rain season supports the analysis that the spatial heterogeneity in canopy at two spatial scales is another reason for differences in H_EC and H_LAS and that canopy has a negative correlation with (H/Rn). An analysis on the influence of the difference in aerodynamic roughness lengths at two spatial scales on sensible heat fluxes shows that the influence is greater in the dry season and smaller in the rain season because the ratio of z_{0m_LAS} to z_{0m_EC} is big in the dry season and is close to 1.0 in the rain season. This study on spatial scales on surface fluxes in the Tibetan Plateau will be helpful in analyzing and understanding its influence on climate.