ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume V-3-2020
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-3-2020, 745–750, 2020
https://doi.org/10.5194/isprs-annals-V-3-2020-745-2020
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-3-2020, 745–750, 2020
https://doi.org/10.5194/isprs-annals-V-3-2020-745-2020

  03 Aug 2020

03 Aug 2020

ASSESSING THE MICRO-SCALE TEMPERATURE-HUMIDITY INDEX (THI) ESTIMATED FROM UNMANNED AERIAL SYSTEMS AND SATELLITE DATA

K. Iizuka1 and Y. Akiyama1,2 K. Iizuka and Y. Akiyama
  • 1Center for Spatial Information Science, The University of Tokyo, Chiba, Japan
  • 2Tokyo City University, Faculty of Architecture and Urban Design / Graduate School of Integrative Science and Engineering, Tokyo, Japan

Keywords: Urban Heat Island, Temperature-Humidity Index, Unmanned Aerial System, Sentinel-2, Random Forest

Abstract. Direct heat and moisture conditions can lead to discomfort for humans and animals and can decrease health performance. The discomfort index or temperature-humidity index (THI) represents an important indicator that measures the heat sensed by humans for different climate conditions. In extreme situations, heatstroke may occur, which in unfortunate cases will lead to death. Many research studies have been conducted on the urban heat island (UHI) phenomenon, although a majority of such work focuses on regional-scale analyses and emphasizes the thermal trend through larger administrative units. Fewer micro-scale analyses have been performed at the local scale to detect the potential area for increased THI within a city. This work seeks to estimate the THI at the micro-scale level by utilizing the thermal camera on-board of unmanned aerial systems (UASs). Thermal information of the surface and visual images are collected by the UAS, while a thermohygrometer is used to collect the air temperature and the relative humidity at the ground surface for ground truth information. Solar radiation and wind exposure modeled from digital surface model (DSM) and normalized difference vegetation index (NDVI) data are used as explanatory variables, and a random forest machine learning method is implemented to model the spatial distribution of the THI. The results and discussion will provide future possibilities for micro-scale analyses of the UHI.