LB 40-334 - Canopy cover and land surface temperature relationship at various scales

Microclimate changes significantly impact habitat quality and ecological patterns. However, it is difficult to measure microclimate in large areas directly with current technologies. Land Surface Temperature (LST) is used as one of the proxies for microclimate. For protecting habitat quality, it is important to identify factors that influence LST at a fine scale. Studies found an inverse relationship between LST and canopy cover using the Landsat remote sensing. However, most of the studies using Landsat have been conducted mainly at 30 m resolution, so there is a limit to grasping the relationship in terms of microclimate in a higher resolution. Moreover, measuring LST through Landsat can be misrepresented in wooded areas. Therefore, to overcome the limitations, we attempted to use Lidar and thermal cameras that can collect high-resolution data. This study aims (1) to determine whether the negative correlation between canopy cover and LST is significant even at a fine scale, and (2) to identify the spatial scale at which this relationship is best explained. We measured the LST, canopy cover data, and terrain information such as DEM at 0.5 m, 1 m, 3 m, 5 m, and 10 m resolution using Back-pack LiDAR, UAV LiDAR, and thermal cameras, at an academic forest located in Gyeonggi Province, South Korea. As a result, the adjusted R-squared value at a 5 m resolution was the highest at 0.47. Moreover, canopy cover and LST showed a negative correlation. These findings were consistent even at large scales. The adjusted R-squared values tended to increase as the resolution decreased, but then decreased at a 10 m resolution. Thus, we found that a 5 m resolution is the most appropriate for analyzing the relationship between LST and canopy cover. Nevertheless, canopy cover and LST showed an insignificant relationship depending on the resolution and the coefficient was relatively low. Adding variables known to affect LST in previous studies could be helpful to improve the low accuracy. In addition, since the ground surface temperature can be influenced by the time of day, considering the measurement time would also be essential.