Date of Degree


Document Type


Degree Name



Earth & Environmental Sciences


Kyle McDonald

Committee Members

Andrew Reinmann

Nir Krakauer

Naresh Devineni


Crop monitoring, Sentinel-1A SAR, ECOSTRESS TIR, MIMICS, DSSAT


Improving crop monitoring, both spatially and temporally, is a key factor in adapting agriculture to the effects of seasonal variability and climate change. Since 2014, numerous space-based remote sensing platforms have been deployed to increase land surface monitoring, with an objective to improve vegetation and crop monitoring. Space-based remote sensing platforms that operate in the thermal infrared (TIR) and microwave wavelengths have proven useful to assess crop conditions. Assessing the ability of TIR and microwave measurements, individually and in combination, offers the ability to reduce measurement gaps and retrieve crop conditions, e.g. growth and health, that can enhance farm management practices.

This dissertation examines the ability to monitor and identify crop conditions both spatially and temporally using the synthetic aperture radar (SAR) imagery, a sensor that transmits microwave radar, and TIR measurements. The analysis focuses on wheat, rice, corn, and vineyard fields located in Yolo and Sonoma County, California, and Long Island New York. To achieve this analysis, SAR imagery from the European Space Agency’s Sentinel-1A (C-band, λ = ~5.5cm) SAR satellite and NASA’s ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS), a sensor that measures TIR emissions (λ = 8 – 14μm) to provide land surface temperature (LST) measurements, were selected to monitor and identify crop conditions.

In Yolo County, California, the ability to identify and measure staple-crop (wheat, rice, and corn) growth and variability was explored using Sentinel-1A SAR with crop simulation and radiometric models. Sentinel-1A SAR results for staple-crops indicated radar backscatter was strongly influenced by crop vegetation structures, i.e. corn leaves and wheat and rice stems, depending on the timing of observation during the growing season. This was verified with a radiative transfer model, which modeled backscatter from each crop-type to understand the salient crop features that influenced SAR response. Modeled backscatter from a wheat, rice, and corn canopy were combined with Sentinel-1A backscatter to produce a SAR-based crop growth index, allowing for the identification of crop fields with high and low growth relative to a modeled backscatter benchmark growth indicator.

For vineyards in Long Island, New York, an analysis was performed to identify Sentinel-1A SAR sensitivity and response to vineyard features, vine canopy growth and moisture content, and soil moisture content, using in situ measurements and radiometric modeling. For vineyards in Sonoma County, the Long Island SAR-vineyard sensitivity analysis was continued and expanded upon by exploring the synergistic capabilities to monitor vineyards with Sentinel-1A SAR and ECOSTRESS LST and evapotranspiration (ET) datasets.

Sentinel-1A SAR results from both vineyard studies indicated SAR backscatter was most sensitive to vine canopy features, vine leafiness and water content, with some sensitive to soil moisture depending on the vine growth stage. This sensitivity response was confirmed by modeling a vine canopy’s interaction with backscatter with a radiative transfer model. In the Sonoma County, California study, Sentinel-1A SAR was sensitive to leafiness and water content, when compared to in situ data and with radiative transfer modeling. ECOSTRESS LST and ET results indicate both datasets can identify vineyard field variation. A Spearman’s rank (Rs) correlation and linear regression analysis was applied between Sentinel-1A backscatter and backscatter ratios to ECOSTRESS LST and ET, producing varying levels of relationships. Overall, Sentinel-1A cross-polarization backscatter, and LST and ET had the strongest relationships in the correlation and regression analysis, but only apparent in the latter due to large temperature differences observed in the vineyard fields. This indicates Sentinel-1A SAR, a ECOSTRESS LST and ET can identify vine growth with vine conditions.

In all, this dissertation concludes SAR imagery can provide important spatial and temporal measurements for wheat, rice, corn, and vineyard fields related to crop growth and moisture conditions. Additionally, the Sonoma study concludes TIR-derived datasets, LST and ET, can provide important information on vine temperature and growth. LST and ET with SAR measurements can be used jointly to assess vineyard conditions, which suggests both measurements types can be extended to other crops to assess their conditions. Further work is needed to assess SAR imagery with TIR measurements but offers the opportunity to expand our crop monitoring efforts with remote sensing observations.

This work is embargoed and will be available for download on Thursday, September 30, 2021

Graduate Center users:
To read this work, log in to your GC ILL account and place a thesis request.

Non-GC Users:
See the GC’s lending policies to learn more.