Impact of Daily Melt and Freeze Patterns on Sea Ice Large Scale Roughness Features Extraction

Hudier, E.; Gosselin, Jean-Sébastien

doi:10.5772/9105

Cited by 1 publication

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Agricultural practices like tillage directly affect surface roughness, influencing radar return. Smooth surfaces act as specular reflectors, deflecting most radar energy and resulting in weak signals, while rough surfaces scatter microwave energy, producing strong and diffuse signals [35]. Co-and cross-polarized waves interact with the surface differently, yielding distinct backscatter amplitudes to the sensor.…”

Section: Introductionmentioning

confidence: 99%

“…While previous studies on frozen states have primarily employed a rigid classification system, i.e., frozen versus thawed states, this work introduces the concept of soil freezing probability, allowing us to use a probabilistic interpretation of soil FT states and study their complex spatio-temporal dynamics in agro-forested landscapes. This complexity arises from various factors that influence radar signals by altering surface roughness, such as spatial variability in soil texture and changes in crop types, along with the intricate interactions among these variables [34][35][36]. The aim of this study encompasses three objectives: First, we analyze the freeze-thaw variability using freezing probability derived from in situ soil temperature at 2 and 10 cm at two study sites over the study periods of 2020-21 and 2021-22.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Sentinel-1-Based Soil Freeze–Thaw Detection in Agro-Forested Areas: A Case Study in Southern Québec, Canada

Taghipourjavi,

Kinnard,

Roy

2024

Remote Sensing

View full text Add to dashboard Cite

Nearly 50 million km2 of global land experiences seasonal transitions from predominantly frozen to thawed conditions, significantly impacting various ecosystems and hydrologic processes. In this study, we assessed the capability to retrieve surface freeze–thaw (FT) conditions using Sentinel-1 synthetic aperture radar (SAR) data time series at two agro-forested study sites, St-Marthe and St-Maurice, in southern Québec, Canada. In total, 18 plots were instrumented to monitor soil temperature and derive soil freezing probabilities at 2 and 10 cm depths during 2020–21 and 2021–22. Three change detection algorithms were tested: backscatter differences (∆σ) derived from thawed reference (Delta), the freeze–thaw index (FTI), and a newly developed exponential freeze–thaw algorithm (EFTA). Various probabilistic mixed models were compared to identify the model and predictor variables that best predicted soil freezing probability. VH polarization backscatter signals processed with the EFTA and used as predictors in a logistic model led to improved predictions of soil freezing probability at 2 cm (Pseudo-R2 = 0.54) compared to other approaches. The EFTA could effectively address the limitations of the Delta algorithm caused by backscatter fluctuations in the shoulder seasons, resulting in more precise estimates of FT events. Furthermore, the inclusion of crop types as plot-level effects within the probabilistic model also slightly improved the soil freezing probability prediction at each monitored plot, with marginal and conditional R2 values of 0.59 and 0.61, respectively. The model accurately classified observed binary ‘frozen’ or ‘thawed’ states with 85.2% accuracy. Strong cross-level interactions were also observed between crop types and the EFTA derived from VH backscatter, indicating that crop type modulated the backscatter response to soil freezing. This study represents the first application of the EFTA and a probabilistic approach to detect frozen soil conditions in agro-forested areas in southern Quebec, Canada.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%