Hyper-temporal remote sensing for digital soil mapping: Characterizing soil-vegetation response to climatic variability

Authors: Maynard, Jonathan; Levi, Matthew

Submitted to: Geoderma
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/22/2016
Publication Date: 1/1/2017
Publication URL: https://handle.nal.usda.gov/10113/5564231
Citation: Maynard, J.J., Levi, M.R. 2017. Hyper-temporal remote sensing for digital soil mapping: Characterizing soil-vegetation response to climatic variability. Geoderma. 285:94-109.

Interpretive Summary: Natural and managed landscapes are being subjected to increasing rates of change that are occurring across a broad range of spatial scales. Central to our understanding of the rate and extent of ecosystem change requires a spatially explicit understanding of the soil resources, since soils modulate ecosystem response to external drivers. Recent technological and methodological advancements in the field of remote sensing are providing new opportunities for utilizing high temporal frequency satellite imagery stacks to predict a range of soil properties. We argue that hyper-temporal remote sensing (RS) (i.e., hundreds of images) can provide novel insights into soil spatial variability by quantifying the temporal response of land surface spectral properties. In this paper we review and synthesize, through the lens of temporal variability, previous research examining the use of RS imagery for DSM. We then developed a conceptual framework for understanding the relationship between vegetation spectral response, which serves as a ‘fingerprint’ of the soil-vegetation-climate relationship, and different soil properties. Finally, we present a case study evaluating the efficacy of the hyper-temporal RS approach for predicting soil texture and coarse fragment classes in a semi-arid ecosystem. Using a 29 year time series of normalized difference vegetation index (NDVI) from Landsat TM data and the support vector machine (SVM) classification algorithm, we effectively modeled both soil texture and coarse fragment classes, where both variables were 60% percent correctly classified (PCC). Results from this study demonstrate the efficacy of the hyper-temporal RS approach in predicting soil properties and highlights how hyper-temporal RS can improve current methods of soil mapping efforts through its ability to characterize subtle changes in RS spectra relating to variation in soil properties.

Technical Abstract: Indices derived from remotely-sensed imagery are commonly used to predict soil properties with digital soil mapping (DSM) techniques. The use of images from single dates or a small number of dates is most common for DSM; however, selection of the appropriate images is complicated by temporal variability in land surface spectral properties. We argue that hyper-temporal remote sensing (RS) (i.e., hundreds of images) can provide novel insights into soil spatial variability by quantifying the temporal response of land surface spectral properties. This temporal response provides a spectral ‘fingerprint’ of the soil-vegetation relationship which is directly related to a range of soil properties. To evaluate the hyper-temporal RS approach, this study first reviewed and synthesized, within the context of temporal variability, previous research that has used RS imagery for DSM. From this analysis we developed a conceptual framework for understanding the soil-vegetation-climate relationship discernable with hyper-temporal RS, which encapsulates both intra- and inter-annual variability. Finally, we demonstrate the utility of this approach by presenting a case study in a semiarid landscape of southeastern Arizona, USA; where surface soil texture and coarse fragment classes were predicted using a 29 year time series of normalized difference vegetation index (NDVI) from Landsat TM data and modeled using support vector machine (SVM) classification. Results from the case study show that SVM classification using hyper-temporal RS imagery was effective in modeling both soil texture and coarse fragment classes, where both variables were 60% percent correctly classified (PCC). In this semiarid ecosystem, variability in precipitation at short time scales (i.e., <6 weeks) was the dominant driver of vegetation spectral variability, with maximum variability corresponding to short-term transitions between climatic states. These short-term transitions between climatic states corresponded to the RS scenes with the highest variable importance in our SMV models, confirming the importance of spectral variability in predicting soil texture and coarse fragment classes. Results from the case study demonstrate the efficacy of the hyper-temporal RS approach in predicting soil properties and highlights how hyper-temporal RS can improve current methods of soil mapping efforts through its ability to characterize subtle changes in RS spectra relating to variation in soil properties.