Abstract
The purpose of this research was to use empirical models to monitor temporal dynamics of soil moisture in a peatland using remotely sensed imagery, and to determine the predictive accuracy of the approach on dates outside the time series through statistically independent validation. A time series of seven Moderate Resolution Imaging Spectroradiometer (MODIS) and Synthetic Aperture Radar (SAR) images were collected along with concurrent field measurements of soil moisture over one growing season, and soil moisture retrieval was tested using Linear Mixed Effects models (LMEs). A single-date airborne Light Detection and Ranging (LiDAR) survey was incorporated into the analysis, along with temporally varying environmental covariates (Drought Code, Time Since Last Rain, Day of Year). LMEs allowed repeated measures to be accounted for at individual sampling sites, as well as soil moisture differences associated with peatland classes. Covariates provided a large amount of explanatory power in models; however, SAR imagery contributed to only a moderate improvement in soil moisture predictions (marginal R2 = 0.07; conditional R2 = 0.7, independently validated R2 = 0.36). The use of LMEs allows for a more accurate characterization of soil moisture as a function of specific measurement sites, peatland classes and measurement dates on model strength and predictive power. For intensively monitored peatlands, SAR data is best analyzed in conjunction with peatland Class (e.g., derived from an ecosystem classification map) to estimate the spatial distribution of surface soil moisture, provided there is a ground-based monitoring network with a sufficiently fine spatial and temporal resolution to fit the LME models.
Highlights
Peatlands are characterized by persistent soil saturation at or near the surface [1] and develop primarily in cool climates [2]
Using the pooled data in linear mixed effects models generally did not lead to strong predictive power for soil moisture
All models resulted in a low marginal R2 but higher conditional R2
Summary
Peatlands are characterized by persistent soil saturation at or near the surface [1] and develop primarily in cool climates [2] They perform many important environmental functions, including the regulation of carbon and water cycling from local to global scales [3,4]. Measurements or simulations of soil moisture and water table depth are key inputs to carbon models [6,7,8] Such models often use point-scale measurements of these parameters, which are extrapolated across peatland landscapes. Measurement or interpolation errors can lead to unrealistic estimates of peatland hydrology and errors in estimates of greenhouse gas outputs [7] These point measurements are not practical across large areas and may not be valid at larger scales or over time [9,10,11]. Information can be captured both at broader spatial scales and repeatedly over time
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
