Date of Award

1996

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Biological & Agricultural Engineering

Abstract

This study relates Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) imagery to field measurements of corn and soybean residue and green crop covers in three Indiana study sites. The primary research objective is to identify the types and amounts of field residue in selected fields in northern Indiana. A major application of this type of data is to provide input into models that focus on soil erosion. A 126-channel subset of 219-channel AVIRIS data was developed for research. Representative spectral curves of baresoils and soils containing varying amounts of corn or soybean residue and green vegetation were analyzed. Two data processing approaches were implemented in this study: (1) simple data transformation and regression analyses: Vegetation and soil indices were applied to selected red, near-infrared, and mid-infrared channels to delineate pixels with mixed composition into their component parts. Pearson analyses were used to estimate the percentages of green and soil cover separately in these study sites. The r-squared values for both features in all study sites were above 0.83. The differentiation between corn and soybean residue was determined from an index in which corn had distinctly higher values than soybeans. The predicted amounts of baresoil, residue, and green vegetation were found within 18% of known values from field measurements; (2) pixel unmixing analysis: Minimal Noise Fraction (MNF) transformation algorithm was applied to the AVIRIS site 1 data. Endmember sampling and pixel unmixing analyses were implemented. The results showed that images derived from pixel unmixing unmixed only spectrally dominant and distinct features. It did not work when two features were spectrally mixed with similar percentages which made the quantitative analysis of this type of application difficult. It is concluded that the first approach, simple yet effective, has the potential to be recognized as an useful tool in predicting and separating crop residue.

Download

Share

COinS