Quickbird satellite and ground-based multispectral data correlations with agronomic parameters of irrigated maize grown in small plots

摘要

Satellite remote sensing has not been practical for agronomic research conducted using small plots due to spatial resolution issues. The objectives were to determine, (1) the potential use of QuickBird high resolution multispectral images for estimating available nitrogen (N) at planting, leaf N content at the tassel (VT) growth stage, and yield (biomass and grain) of irrigated maize (Zea mays L.) grown in research size plots and (2), if image digital number (DN) values were adequate for assessing these agronomic parameters. Conventionally-tilled plots 21.3 m x 21.6 m, 10.7 m x 43.3 m, and 10.7 m x 21.6 m were used. QuickBird images (2.5 m pixel resolution) were obtained at the VT growth stage for two growing seasons along with ground-based reflectance for one growing season. Multispectral data [blue (450-520), green (520-600), red (630-690), and near-infrared (760900 nm)] was acquired by both sources. Relative green (G) waveband DN had the highest correlations with available N at planting (r(2)=0.91), relative grain yield (r(2)=0.81) and relative total biomass (r(2)=0.59) than any of the other indices investigated. Relative spectral indices containing G and near-infrared (NIR) wavebands, i.e., the green normalised difference vegetation index (GNDVI), the NIR/G ratio and the chlorophyll index ((NIR/G)-1) produced good correlations with leaf N content at VT (r(2) > 0.7). Regression parameters (slopes and intercepts) of the normalised (NIR/G)-1 index calculated from satellite DN values and ground-based reflectance versus leaf N content were not significantly different (alpha=0.01). Results indicate that QuickBird multispectral images show promise for estimating agronomic parameters from small plot research. Additional research is required to expand the results to more maize growth stages and other crops.