Date of Award

Spring 1-1-2019

Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

Albin Gasiewski

Second Advisor

Dejan Filipovic

Third Advisor

Dimitra Psychogiou

Fourth Advisor

Brian Argrow

Fifth Advisor

Venkat Lakshmi


An L-band Lobe Differencing Correlation Radiometer (LDCR) was designed, developed and demonstrated for high spatial resolution soil moisture mapping from a small unmanned aerial system (sUAS). A lightweight lobe-differencing antenna was designed as a 2x2 rectangular L-band microstrip array with effective use of styrofoam blocks for frequency tuning and mechanical stability. An optimal vertical separation distance was identified that minimizes mutual coupling and provides maximum main-to-back lobe ratio. A method of antenna frequency tuning using dielectric overlay material is presented to correct for the dielectric impact of the sUAS fuselage. The LDCR RF front end design and performance are detailed. The LDCR was calibrated using both pre-flight lab test data and in-flight data over a calm pond to determined its sensitivities and offset which are compared with analytical values. Radio frequency interference (RFI) from the sUAS platform was observed prior to platform integration prompting RFI mitigation to be performed upon integration of the radiometer into the platform. Flight tests of the LDCR on the Black Swift Technology's Tempest sUAS were performed at the Canton, Oklahoma Soilscape site in September 2015 and Irrigation Research Foundation (IRF) in Yuma Colorado in June 2016. Based on a soil vegetation radiative transfer (SVRT) model for the LDCR, the LDCR TA measurements observed on the sUAS sampling grid were mapped to estimated volumetric soil moisture (VSM) on a regular user-defined mapping grid. The estimator used a new full-domain linear minimum mean square error (LMMSE) VSM retrieval algorithm, along with LDCR infrared physical temperature TP measurements, Landsat-based vegetation water content (VWC) maps, and soil texture information. The LDCR error budget for this algorithm was developed. The initially retrieved VSM data are favorably compared with in-situ measured VSM data and irrigation records.