A Time-Series-Based Approach to Modeling Planetary Surface Change

Abstract:

Dust lifting and deposition represent the most dynamic surface change processes across much of the martian surface. The interaction of dust reservoirs with the atmosphere represents the surface branch of the global dust cycle. Depletion and replenishment of these reservoirs may be an important factor in the interannual variability of the martian atmosphere, including the generation of large regional and global-scale storms. Orbiting spacecraft equipped with wide-angle, multispectral imaging cameras have provided near-continuous monitoring of the atmosphere and surface with varying spectral coverage and spatial resolution since the arrival of Mars Global Surveyor (MGS) in the mid-1990s. More recently, the Mars Reconnaissance Orbiter (MRO) Mars Color Imager (MARCI) instrument has provided near-daily global coverage since 2006, and is expected to continue operations through this decade.

On Earth, dynamic surface changes typically represent the action of very different processes than those occurring on Mars. Urbanization, forest harvest and regrowth, agriculture, and other natural and human-driven processes are the dominant drivers of change. Multi-date comparisons performed on Earth satellite images often focus on specific spectral difference ratios (such as vegetative indices) that are uniquely terrestrial in application. Increasingly, however, the availability of large-scale, pre-processed data products, such as U.S. Geological Survey (USGS) Landsat Analysis Ready Data (ARD), has driven the rise of time series data analyses. One such algorithm is being implemented by the USGS Earth Resources Observation and Science (EROS) center at a national scale as part of the Land Change Monitoring, Assessment, and Projection (LCMAP) initiative. This approach focuses on pixel-based harmonic time-series spectral modeling to characterize the seasonal phenology of the surface and identify significant deviations that represent abrupt change. This method is not fundamentally tied to either vegetative landscapes or to terrestrial datasets specifically. Application of a similar algorithm to Mars would allow global characterization of seasonal and interannual variations in surface dust cover, and rapid identification of dust removal, with the potential for important advances in martian climate studies.