Remote sensing, normalized difference vegetation index (NDVI), and crop yield forecasting

The thesis explored the feasibility of using remotely sensed image and its derived products, Normalized Difference Vegetation Index (NDVI), to assess and quantify corn and soybean yield potential. Fixed-effect panel and ordinary least squares NDVI regression models were developed for different level of spatial aggregation. Through the regression analysis, the thesis identified the relationship between the accumulation of crops’ “greenness” over the growing season and the final crops yield. The ultimate goal of the thesis is to examine whether the NDVI model can produce accurate and timely yield forecasts. Due to the unique features of the spatial data (e.g. global coverage, frequent repeat cycle and etc.), the model can provide significant value to developing countries where the meteorological network is scarce and official crop production estimates are either inaccurate or nonexistent. Therefore, to evaluate the NDIV model’s predictive power, the model’s out-of-sample forecasts were compared to the predictions of a weather-based regression model (modified Thompson model) as well as August, September, and October USDA estimates.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Agricultural & Applied Econ

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2014

Author dc:creator

• Lv, Xijie

Contributors dc:contributor

• Paulson, Nicholas D.

• Remote Sensing
• Normalized Difference Vegetation Index (NDVI)
• Crop Yield Assessment
• Forecasting

Statement dc:rights

• Copyright 2013 Xijie (CJ) Lv

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/46590

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/46590