Title: The Influence of Sun Position and Clouds on Reflectance and Vegetation Indices of Greenhouse-Grown Corn Authors
|Souza, Edwardo - UNIOESTE, BRAZIL|
|Scharf, Peter - UNIVERSITY OF MISSOURI|
Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: August 10, 2009
Publication Date: February 1, 2010
Citation: Souza, E.G., Scharf, P.C., Sudduth, K.A. 2010. The Influence of Sun Position and Clouds on Reflectance and Vegetation Indices of Greenhouse-Grown Corn. Agronomy Journal. 102:734-744. Interpretive Summary: For most grain crops, farmers apply some form of nitrogen (N) to help meet the crop’s nutrient needs. While current practices generally apply one rate of N over a whole field, or even an entire farm, research shows that crop N need often differs dramatically within fields. Consequently, in places where a single N rate exceeds crop need, the left-over N may be lost to the environment. In places where a single rate is less than crop N need, yield will be less than optimal. One promising approach for better matching N need and application is within-field sensing of the reflectance characteristics of the crop and subsequent N application based on this sensed information. However, because the intensity of sunlight varies with time of day and sky conditions, data from sunlight-based sensors also varies, even if the crop characteristics are constant. One way to overcome this obstacle is to use sensors that include their own light source. Another approach, described in this paper, is to compensate for differences in sunlight as a function of sun position and cloudiness. We developed compensation methods applicable to within-field on-the-go sensing and tested these methods on corn. We were able to remove one-third to one-half of the sunlight-induced variability in the sensor data, thus making it more useful for N application decisions. These findings will be useful to those using sunlight-based sensors for N management or other crop reflectance measurements, allowing improved interpretation of the sensor data.
Technical Abstract: The reflectance characteristics of plants and plant canopies far from solar noon or with cloudy skies are not well known. This is an obstacle to making real-time variable-rate N fertilizer applications based on canopy reflectance because such a system must work under cloudy skies and at all times of day. Our objective for this project was to develop spectral radiometer reflectance corrections for variations in incoming sunlight so that the same reflectance reading would be obtained (and the same N recommendation made) for the same plants regardless of time of day or cloud conditions. Spectral radiometers were mounted in a stationary position about twenty-five centimeters above the corn canopy. Readings were taken from morning until night over several days with a range of sky conditions (sunny, overcast, and partly cloudy). Experiments were done in the field on greenhouse-grown corn ranging from V10 to R2 growth stages. Sun angle, time of day, and cloud cover all influenced reflectance measured from the corn canopy. When regression models were applied to correct reflectance values to reference conditions for these variables, coefficients of variation were reduced by 29 to 56% for vegetation indexes and by 43 to 56% for reflectance values. The NIR/green ratio and the Green Normalized Difference Vegetation Index (GNDVI) were the indices most sensitive to N deficiency among six analyzed indices.