Main content area

Assessing corn water stress using spectral reflectance

DeJonge, Kendall C., Mefford, Brenna S., Chávez, José L.
International journal of remote sensing 2016 v.37 no.10 pp. 2294-2312
Zea mays, corn, crop yield, data collection, growers, growing season, guidelines, irrigation management, microirrigation, normalized difference vegetation index, reflectance, remote sensing, rhizosphere, soil water, soil water balance, vegetation, water stress, Colorado
Multiple remote-sensing techniques have been developed to identify crop-water stress; however, some methods may be difficult for farmers to apply. If spectral reflectance data can be used to monitor crop-water stress, growers could use this information as a quick low-cost guideline for irrigation management, thus helping save water by preventing over-irrigating and achieving desired crop yields. Data was collected in the 2013 growing season near Greeley, Colorado, where drip irrigation was used to irrigate 12 corn (Zea mays L.) treatments with varying water-deficit levels. Ground-based multispectral data were collected and three different vegetation indices were evaluated. These included the normalized difference vegetation index (NDVI), the optimized soil-adjusted vegetation index (OSAVI), and the Green normalized difference vegetation index (GNDVI). The three vegetation indices were compared to water stress as indicated by the stress coefficient (K ₛ), and water deficit in the root zone was calculated using a soil water balance. To compare the indices to K ₛ, vegetation ratios were developed from vegetation indices in the process of normalization. Vegetation ratios are defined as the non-stressed vegetation index divided by the stressed vegetation index. Results showed that vegetation ratios were sensitive to water stress as indicated by the good coefficient of determination (R ² > 0.46) values and low root mean square error (RMSE < 0.076) values when compared to K ₛ. To use spectral reflectance to manage crop-water stress, an example irrigation trigger point of 0.93 for the vegetation ratios was determined for a 10–12% loss in yield. These results were validated using data collected from a different field. The performance of the vegetation ratio approach was better than when applied to the main field giving higher goodness of fit values (R ² > 0.63), and lower error values (RMSE < 0.043) between K ₛ and the vegetation indices.