The Capability of GLOBE Data to Correct for Atmospheric Differences in Land Surface Temperature
Abstract
Surface temperature is a quantitative and continuous phenomenon, understood easily
when using data collected remotely by satellites. Surface temperature has been studied to
further investigate the warming of urban climates, also known as urban heat islands, which can
impact urban planning, public health, pollution levels, and energy consumption. However, the
full potential of remotely sensed images is limited, when analyzing land surface temperature,
due to the daunting task of correcting for changes in the atmosphere. To help address this,
Landsat 8 employed a new thermal infrared sensor which records land surface temperature in
two wavelengths. With two bands in the infrared region, a split window algorithm (SWA) was
applied to correct for atmospheric difference. Split window algorithms are commonly practiced
when correcting for atmospheric differences observed by other satellites with two bands in the
thermal infrared region. The purpose of this research was to use in situ surface temperature
measurements from NASA’s ground observation program, the Global Learning and
Observations to Benefit the Environment (GLOBE), to derive the correcting coefficients for use
in the split window algorithm. The GLOBE database provided land surface temperature data
that coincided with Landsat 8 overpasses to create a global model for correcting Landsat
scenes across the world. While it is possible to derive the coefficients from the GLOBE data, the
results indicate that more GLOBE data needs to be collected to create a model with complete
statistically significant coefficients.