Skip to main content
ARS Home » Plains Area » Las Cruces, New Mexico » Range Management Research » Research » Publications at this Location » Publication #321525


Location: Range Management Research

Title: A hyper-temporal remote sensing protocol for detecting ecosystem disturbance, classifying ecological state, and assessing soil resilience

item Maynard, Jonathan
item Karl, Jason

Submitted to: Soil Science Society of America Annual Meeting
Publication Type: Abstract Only
Publication Acceptance Date: 8/21/2015
Publication Date: 11/16/2015
Citation: Maynard, J.J., Karl, J.W. 2015. A hyper-temporal remote sensing protocol for detecting ecosystem disturbance, classifying ecological state, and assessing soil resilience [abstract]. 2015 Soil Science Society of America Annual Meeting. November 15-18, 2015, Minneapolis, MN. p. 39-17.

Interpretive Summary:

Technical Abstract: Hyper-temporal remote sensing is capable of detecting detailed information on vegetation dynamics relating to plant functional types (PFT), a useful proxy for estimating soil physical and chemical properties. A central concept of PFT is that plant morphological and physiological adaptations are linked in predictable ways by resource limitation, responses to disturbance, and edaphic factors. Consequently, the main objectives of this study were to evaluate the ability of hyper-temporal remotely sensed imagery to (i) classify PFT relating to different ecological states, (ii) quantify the magnitude and timing of significant ecosystem changes, and (iii) assess the dominant soil properties that impart ecosystem resilience within a semi-arid rangeland ecosystem. The modified soil adjusted vegetation index (MSAVI), derived from Landsat imagery, was compiled for the period between 2000 and 2012 (16-day frequency), resulting in 269 observations per 30-m pixel. Hierarchical cluster analysis was performed on the hyper-temporal image stack, delineating areas with similar ecological states based on temporal variability within each time-series. The average temporal signal from each cluster was then decomposed into seasonal and long-term trend components using the Breaks For Additive Seasonal and Trend (BFAST) model, which further identified the direction, magnitude and timing of significant deviations from modeled trends. Significant breaks, in combination with ancillary environmental data, were used to evaluate the resistance and resilience of each ecological state. Soil properties maps (0-50 cm), created from gridded SSURGO data, provided explanatory variables in an ordination and variance partitioning analysis to determine which key soil properties correspond to identified patterns of ecosystem resilience. Results from this study showed that physical soil properties (e.g., texture, water holding capacity) were the most significant in explaining resilience patterns. Furthermore, this study demonstrates the power and utility of this approach in delineating ecological states, interpreting temporal drivers of change, and identifying edaphic controls on ecosystem resilience.