Rakhesh Devadas

Interaction of Nitrogen Application and Stripe Rust Infection in Wheat Using In-situ Proximal and Remote Sensing Techniques

RakheshDevadas sq
University
University of New England
Supervisor (Academic)
Assoc Prof David Lamb & Dr David Backhouse, UNE
Supervisor (Industry)
Dr Steven Simpfendorfer, DPI NSW
Projects
mysite
Employment
Data Manager/Researcher, University of Technology, Sydney
Thesis Abstract

The project dealt with modelling the interaction of nitrogen nutrition and stripe rust (yellow rust) incidence in wheat using spectral reflectance characteristics at different spatial scales as observed by ground based sensors, airborne and satellite data.

Experimental plots, with different levels of N application, variety and seed treatment for stripe rust disease, were set up in crop seasons 2006 and 2007. Temporal ground based multispectral data were collected using Crop Circle ACS-210 (Holland Scientific Inc., NE, USA) and the GreenSeeker model 505 (Ntech Industries Inc., CA, USA). Hyperspectral data were collected using USB 2000 (Ocean Optics, FL, USA). This ground based data were analysed in relation to airborne data collected using an airborne multispectral sensor, UNEBiRD (UNE, Armidale). Multispectral and hyperspectral vegetation indices (VIs) derived from the two years of data were analysed in relation to the occurrence of N deficiency, disease incidence, LAI, chlorophyll content, biomass and yield in wheat. Further, applicability of these VI based models at a higher spatial scale was examined employing multispectral (Landsat 5TM ) and hyperspectral (EO1 Hyperion) satellite data acquired over commercial wheat paddocks in northern NSW, Australia.

Analysis of agronomic data confirmed the expected outcomes of a positive correlation between N application and yield up to a certain rate of N application, with further addition of N causing yield to plateau or subsequently decrease. This study also confirmed that there was significant positive correlations between N application and stripe rust severity.

Temporal Normalised Difference Vegetation Index (NDVI) data derived from ground based multispectral sensors were found to be highly effective in modelling LAI and biomass generation. NDVI data collected towards the peak vegetative growth phase were observed to be critical for yield modelling in disease free wheat crops. However, NDVI measurements carried out after the peak vegetation phase were found to increase the accuracy of yield modelling where of the crop was infected with stripe rust. Both N deficiency and stripe rust severity showed highly significant negative correlations with multispectral NDVI values which made separation of N deficiency from disease occurrence difficult using NDVI measurements. Also, it was inferred that NDVI data could capture variations in N deficiency/nutrition more efficiently than that of stripe rust severity.

Hyperspectral data analysis indicated that VI utilizing the changes in leaf pigment concentration characterised by the reflectance pattern in the 530-550 nm waveband, was superior in the estimation of different levels of stripe rust incidence. Conversely, VIs capturing the changes in reflectance in the near infrared (NIR) region (705-750 nm) was observed to be the best indicator of N deficiency. The contrasting behaviour of these VIs, especially Physiological Reflectance Index (PhRI) and Leaf and Canopy Chlorophyll Index (LCCI), make these indices potential tools for discrimination and modelling of stripe rust infection and N deficiency when applied in a sequence. 

VI derived from ground based, airborne and satellite sensors showed strong correlations, which indicated the possibility of utilizing spectral models at a higher spatial scale. However, this correlation declined consistently with decreasing spatial resolution of remote sensing data. This NDVI distortion resulting from changing sensor-target distances caused systematic underestimation of crop yield. However, study demonstrated that the prediction accuracy could be improved by applying a simple empirical conversion equation to convert at-sensor NDVI (Landsat 5 TM) to effective on-ground NDVI using near-coincident on-ground NDVI measurements.