PhD Research
L-Band InSAR global feasibility map for SWE retrieval
Mountain snowpacks provide vital water resources for communities in the western U.S. (WUS), but high variability creates a challenging problem for accurate measurement of snow water equivalent (SWE) from remote sensing platforms. Studies using repeat airborne L-band (~25 cm wavelength) Interferometric Synthetic Aperture Radar (InSAR) have demonstrated sensitivity to forest cover fraction (FCF), liquid water content, and incidence angle. We use these factors to map feasibility of L-band InSAR for ΔSWE retrievals in major mountain ecoregions of WUS. We found feasibility declines from ~65% on 1 February, to 58% on 1 March, and 30% on 1 April, corresponding to 73%, 70%, and 49% of total SWE volume, respectively. Thus, these feasibility maps provide groundwork for future InSAR snow studies using satellite data from missions such as NISAR to refine work based on regional conditions and hence improve ΔSWE retrievals globally. Link to the published article
L-Band InSAR Coherence Controls in a Mountain Snowpack
L-band Interferometric Synthetic Aperture Radar (InSAR) is a powerful technique to measure snow water equivalent (SWE) from space. However, for precise SWE estimation, the two SAR acquisitions need to be coherent. Various factors, including the presence of dense vegetation, liquid water in the snowpack, and certain radar properties such as the incidence angle, can lead to systematic decorrelation between the two SAR images. In this study, we explore the impacts of these three major factors on InSAR coherence for three different pairs (12–19 February, 19–26 February, and 12–26 February 2020). This study builds on the UAVSAR dataset from the NASA SnowEx 2020 campaign in the Jemez mountains of New Mexico by integrating liquid water content (LWC) simulations using a physically based and spatially distributed SnowModel to examine interconnections between snowpack LWC, canopy cover, and radar’s local incidence angle. We model these three factors using ordinary least squares regression and linear mixed-effects models with random slopes, grouped by elevation, slope, and northness. Canopy cover and local incidence angle consistently served as negative predictors and together explained ~41% of the coherence variation, whereas LWC was a weak and temporally unstable predictor which contributed minimally to the explained variance. Incorporating terrain-dependent predictor effects significantly improved the model, with the conditional R2 reaching about 64% of the explainable variance for the elevation by northness grouping. This suggests that vegetation and geometric influences are largely conditioned by topographic controls. With the NISAR mission now operational and providing globally consistent L-band InSAR data, these findings establish a physically grounded basis for understanding coherence in complex mountain snowpack terrains. (Manuscript in preparation)
Estimate the spatial variability of GPR derived snow density during NASA SnowEx 2023 Alaska campaign
Arctic tundra and boreal forest hold over half of the terrestrial snow cover, yet the spatial variability of snow properties in these environments remains difficult to quantify, limiting the accuracy of snow water equivalent (SWE) estimates and snowmelt runoff forecasts. While snow depth variability has been studied extensively, far less attention has been given to the spatial structure of snow density, which wind redistribution alters alongside depth in tundra environments. This study will estimate the spatial variability of snow density derived from ground penetrating radar (GPR) measurements collected during the NASA SnowEx 2023 Alaska campaign at two arctic tundra sites (ACP, UKT) and three boreal forest sites (BCEF, CPCRW, FLCF). GPR two-way travel times from snowmobile-towed spiral surveys will be combined with coincident ground-based snow depth measurements to compute radar velocities, relative permittivities, and snow densities using established petrophysical relationships. Ground-based depths are preferred over LiDAR-derived depths to avoid biases introduced by mismatched acquisition dates, GPR footprint size, and GNSS positioning uncertainty. Semivariograms will be computed at a 3 m lag spacing to estimate correlation lengths of density and SWE within spirals, between spirals, and between sites, spanning point to regional scales. Differences in spatial patterns between snow classes will be tested with the Wilcoxon rank-sum test. We expect wind-driven redistribution into topographic depressions and lee slopes to dominate density variability in the tundra, while canopy interception controls variability in the boreal forest. These results will characterize how snow density heterogeneity differs between the two dominant high-latitude snow classes, informing sampling design and distributed SWE estimation. (Manuscript in preparation)
Masters Research
Cover crop effects on X-ray computed tomography–derived soil pore characteristics
This study aimed to compare the effects of cover crops (CC) and no cover crops (NC) on soil macropore characteristics in strip-tillage cotton fields. Using high-resolution X-ray CT scanning, the results showed that cover crops increased soil porosity and pore number density in the topsoil. Additionally, deeper subsurface layers under CC had higher connection probability, indicating potential influence on subsurface flow pathways. These findings suggest that cover crop roots play a significant role in shaping soil pore structures, which can impact water and contaminant transport. Link to the published article
Characterization of soil pores in strip-tilled and conventionally-tilled soil using X-ray computed tomography
This study investigated the effects of conventional tillage (CT) vs. strip tillage (ST) on soil pore characteristics across two seasons. Soil cores collected from cotton fields were analyzed using X-ray computed tomography. The results showed that ST had significantly higher macroporosity, network density, and pore connectivity compared to CT in the first season, likely due to less soil disturbance. However, in both tillage systems, pore properties decreased significantly in the second season due to soil reconsolidation from rainfall. The findings highlight how tillage practices and seasonal changes influence soil pore morphology and its potential impact on contaminant transport. Link to the published article
Effect of Image Resolution and Soil Core Size on Soil Pore Characteristics
This study aimed to compare the effects of CT scanning resolution and soil core size on detected soil pore properties. Cylindrical soil cores of different diameters (76 mm and 150 mm) were collected from a cotton field under conventional and strip tillage in two seasons. Results showed that higher resolution scanning revealed more isolated pores with greater anisotropy, while smaller core diameters detected fewer pores but had greater pore connectivity. Significant differences were observed mainly in conventional tillage cores from season 2. The findings highlight the tradeoff between resolution, field of view, and core size, emphasizing the importance of aligning sampling methods with research objectives.Link to the published article