Correlation of NDVI calculated with GreenSeeker and VNIR spectroscopy

Optical sensors are recognized as non-destructive and cost effective methods to estimate different vegetation Indices. The objective of this study was to evaluate the reliability of normalized difference vegetative index (NDVI) values calculated by the GreenSeeker (GS) hand held sensor (Trimble, USA) as an indirect indicator of tomato and cucumber Nitrogen (N) status, and compare to the Visible-NIR (VNIR) spectra. NDVI index is commonly used to differentiate plant properties such as chlorophyll, biomass, and plant nutrition [1]. A reflectance VNIR system operating in the range 300-1100 nm was applied to evaluate performance of GS. The experiment was conducted in spring 2017 on tomato and cucumber planted in greenhouse which supplemented with five treatments of 46% Urea controlled release fertilizer (CRF). NDVI readings were recorded every weeks during the growth seasons and spectra of leaves earned in two stages (After the first and last repeated fertilization) like [2]. Sensors values were compared to actual leaf N concentration with Kjeldahl measurement. Results obtained by VNIR spectroscopy combined with chemometric analysis for the determination of N content. Spectra were analyzed by The Unscrambler software (version 10X, Camo, Norway) in order to build Partial least square (PLS) regression models able to estimate NDVI parameter of tomato and cucumber. Spectral data were normalized by Standard Normal Variate (SNV) and pretreated using a derivative transform calculation (Savitzky–Golaymethod) by 1st derivative [3]. Spectral data were elaborated by leave one-out cross validation (LOOCV) of PLS regression technique. High correlation models between spectral data and NDVI of GS and also Kjeldahl method were obtained for the N concentration (0.75 R2 0.89 and 0.81 R2 0.90 for spectra-GS and spectra- Kjeldahl respectively). The GS readings of tomato’s leaves were less correlated with leaf N concentration at early stages of establishment due to plants small size and background noise compared to wide leaves of cucumber