Publication Date

2026

Document Type

Dissertation/Thesis

First Advisor

Gensini, Vittorio A.

Degree Name

Ph.D. (Doctor of Philosophy)

Legacy Department

Department of Geographic and Atmospheric Sciences

Abstract

United States agriculture faces increasing pressure of sustaining food production amid global population growth while increasing weather extremes amplify risk and vulnerability across the food system. This dissertation evaluates historical and future projected changes in crop conditions across ten major field crops: barley (Hordeum vulgare L.), corn (Zea mays L.), cotton (Gossypium hirsutum L.), oats (Avena sativa L.), peanuts (Arachis hypogaea L.), rice (Oryza sativa L.), sorghum (Sorghum bicolor L.), soybeans (Glycine max L.), and spring and winter wheat (Triticum aestivum L.). First, a multi-decadal comparison of the USDA crop condition index and satellite-derived vegetation health indices reveal that vegetation health metrics contain strong explanatory power of yield during early and mid-summer growing season growing stages, whereas the crop condition index maintains or strengthens its association with yield through reproduction and maturation stages. These complementary dynamics support a blended monitoring framework that leverages both subjective crop assessments and satellite-based products, but the subjective USDA data highlight the agronomic value embedded in expert crop assessments, which integrates field-level context beyond what satellites alone can resolve. Second, historical analyses using the USDA crop condition index reveal that conditions deteriorated and became more variable across portions of the Great Plains since the mid-1980s, while improvements and declining variability are observed in parts of the southeastern and southwestern United States. Spatial differences in mean conditions and interannual variability reflect regional climate, crop selection, and management practices, underscoring the role of adaptation and technological advancement in sustaining yield gains. Finally, state-, month-, and crop-specific gradient-boosted machine learning models trained on precipitation, temperature, water balance, and soil moisture predictors are applied to high-resolution, convective-permitting regional climate simulations to quantify crop condition trajectories through the twenty-first century. Results indicate increasingly widespread and spatially coherent late-season condition declines under higher emissions, alongside increasing interannual variability during yield-sensitive phenological periods. Collectively, these findings establish a comprehensive observational-to-projection framework for understanding how crop conditions respond to climate stressors and highlight the need for regionally targeted adaptation strategies to maintain agricultural productivity.

Extent

163 pages

Language

en

Publisher

Northern Illinois University

Rights Statement

In Copyright

Rights Statement 2

NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.

Media Type

Text

Available for download on Wednesday, June 02, 2027

Share

COinS