Date of this Version
Spatial irrigation management has been steadily advancing over the last several years. A current issue with managing irrigation spatially on sub-field scale is the inability to readily collect the spatial field data necessary to properly manage irrigation. Multispectral and thermal infrared imagery used in informing irrigation management decisions was previously collected by satellite and manned aircraft remote sensing platforms. These remote sensing platforms pose issues concerning economic feasibility, revisit intervals, and weather factors that inhibit the collection of data. Recent developments in unmanned aerial systems, which provide an additional means of collecting multispectral and thermal infrared data, have the potential to provide supplemental data during periods of missing satellite data or to completely replace satellite and manned aircraft remote sensing platforms. As unmanned aerial system remote sensing platforms are a relatively new technology, there are uncertainties regarding how these systems compare to previous and more well-known remote sensing platforms. Some of these uncertainties include how to properly collect, process, and calibrate data acquired by these systems so that the end products are accurate and can by used in scientific applications. This work evaluated two different unmanned aerial systems with integrated multispectral and thermal infrared cameras to determine the best methods of collecting, processing, and calibrating data. Three different multispectral image calibration methods were evaluated and compared against Landsat satellite reflectance products and ground-based reflectance tarps. The thermal infrared image calibration consisted of correcting for emissivity and atmospheric effects, and was compared to in-field infrared thermometers. Relationships for estimating maize leaf area index, crop height, and fraction of vegetation cover were redefined and evaluated based on various vegetation indices derived from the unmanned aerial system calibrated multispectral imagery. This work also addressed some of the challenges and obstacles related to deploying unmanned aerial systems for remote sensing in agricultural applications.
Advisors: Wayne E. Woldt and Christopher M.U. Neale