Physically Constrained Spatiotemporal Modeling of Remotely Sensed Land Surface Temperature

Gavin Collins, Brigham Young University

Session 2

Description

Satellite remote-sensing is often used to collect important atmospheric and geophysical data that provide insight into spatial and temporal climate variability over large regions of the earth, at high spatial resolutions. Common issues surrounding such data include missing information in space due to cloud cover at the time of a satellite passing and large blocks of time for which measurements are not available due to the infrequent passing of polar-orbiting satellites. While many methods are available to predict missing data in space and time, in the case of land surface temperature (LST) data, these approaches generally ignore the temporal pattern called the "diurnal cycle" which physically constrains temperatures to peak in the early afternoon and reach a minimum at sunrise. In order to construct a complete, physically justifiable remotely sensed dataset, we parameterize the diurnal cycle into a functional form with unknown spatiotemporal parameters. Using multi-resolution basis functions, we estimate these unknown parameters from sparse satellite observations to obtain physically constrained estimations of LST. The methodology is demonstrated using two remote sensing datasets of LST in Houston, TX and Phoenix, AZ USA, collected by NASA’s Aqua and Terra satellites.

 
May 18th, 9:10 AM

Physically Constrained Spatiotemporal Modeling of Remotely Sensed Land Surface Temperature

Orbital ATK Conference Center

Satellite remote-sensing is often used to collect important atmospheric and geophysical data that provide insight into spatial and temporal climate variability over large regions of the earth, at high spatial resolutions. Common issues surrounding such data include missing information in space due to cloud cover at the time of a satellite passing and large blocks of time for which measurements are not available due to the infrequent passing of polar-orbiting satellites. While many methods are available to predict missing data in space and time, in the case of land surface temperature (LST) data, these approaches generally ignore the temporal pattern called the "diurnal cycle" which physically constrains temperatures to peak in the early afternoon and reach a minimum at sunrise. In order to construct a complete, physically justifiable remotely sensed dataset, we parameterize the diurnal cycle into a functional form with unknown spatiotemporal parameters. Using multi-resolution basis functions, we estimate these unknown parameters from sparse satellite observations to obtain physically constrained estimations of LST. The methodology is demonstrated using two remote sensing datasets of LST in Houston, TX and Phoenix, AZ USA, collected by NASA’s Aqua and Terra satellites.