Authors: Peng Fu*, University of Illinois
Topics: Remote Sensing, Climatology and Meteorology, Environment
Keywords: land surface temperature, urban canopy model, cloud contamination, WRF, thermal infrared
Session Type: Paper
Start / End Time: 9:55 AM / 11:35 AM
Room: Taylor, Marriott, Mezzanine Level
Presentation File: No File Uploaded
Remotely sensed land surface temperature (LST) has been widely utilized for quantifying surface energy and carbon fluxes at different scales. Although a series of algorithms have been developed to estimate LSTs from satellite thermal infrared (TIR) data, the retrieval of LSTs under cloudy conditions has received less attention. The existing techniques, such as those based on passive microwave measurements and surface energy balance, for estimating LSTs under cloudy skies may not fulfil the needs for a spatially and temporally consistent LST dataset. To this end, this study synergistically used the coupled WRF/UCM system and the random forest (RF) regression technique to estimate LSTs for overcast moments. With the Baltimore-Washington metropolitan region used as the study site, the WRF/UCM simulations (LSTs) were conducted with appropriate configurations. The MODIS LST images were downloaded and used to test the developed approach. The utilization of the RF regression technique for estimating LSTs under cloudy conditions from a partially cloud-contaminated LST image greatly reduced the RMSE value to 1.7 K without compromising the correlation coefficient value. For a fully cloud-contaminated LST image, LSTs could still be reconstructed by using neighboring images of a target date. The results suggested that the RF regression technique could provide a relatively smoothed LST image with the correlation coefficient and RMSE values of 0.74 and 2.1 K, respectively. Further research efforts can be made to understand the impacts of the temporal distance between neighboring images and the target image on the performance of the developed approach.