Spectrum analysis of hyperspectral red edge position to predict rice biophysical parameters and grain weight

Rice canopy hyperspectral reflectance data were perpetually measured to analyze the red edge position (REP) corresponded to grain weight during growing season in Java Island, Indonesia. Reflectance measurements were conducted concurrently with biomass sampling of three rice cultivars and four nitrogen application levels (N level) whose were involved in this study. The potential of narrow bands in the red edge region were investigated to predict nitrogen content (N content), plant chlorophyll index (SPAD value) and grain weight with applying the inverted Gaussian (IG) and modified linear extrapolation (LE) methods. Spectral data was smoothed into five nm interval. The first derivative reflectance that derived from reflectance data subsequently used in analysis of REP. The REP moved to longer wavelength until close to heading stage and otherwise shifted to shorter wavelength after heading stage. The correlation REP-IG (R2=0.849) was stronger than REP-LE (R2=0.767) attributed to N content, while the correlation REP with SPAD value, REP-LE (R2=0.800) demonstrated slightly higher than that of REP-IG (R2=0.731) . Subsequently, REP-LE represented slightly more significant correlation with grain weight (R2= 0.884) when compared to REP-IG (R2=0.847) . The REP shift can act as quick and precise prediction for biophysical parameters and grain weight status. Its prediction capability can support crop farming management.

[1]  Compton J. Tucker,et al.  Spectral estimation of grass canopy variables , 1977 .

[2]  F. T. Turner,et al.  Chlorophyll Meter to Predict Nitrogen Topdress Requirement for Semidwarf Rice , 1991 .

[3]  Tsuyoshi Akiyama,et al.  Estimating grain yield of maturing rice canopies using high spectral resolution reflectance measurements , 1991 .

[4]  R. Martin,et al.  Spectral reflectance patterns of flooded rice , 1986 .

[5]  Rong-Kuen Chen,et al.  Modeling Rice Growth with Hyperspectral Reflectance Data , 2004 .

[6]  W. Collins,et al.  Remote sensing of crop type and maturity , 1978 .

[7]  William J. Collins,et al.  Remote detection of metal anomalies on Pilot Mountain, Randolph County, North Carolina , 1983 .

[8]  D. Horler,et al.  The red edge of plant leaf reflectance , 1983 .

[9]  Tsuyoshi Akiyama,et al.  Optimal visible and near-infrared waveband used in hyperspectral indices to predict crop variables of rice , 2008 .

[10]  Jingfeng Huang,et al.  Red edge parameters as indicators of rice nitrogen levels , 2003, SPIE Asia-Pacific Remote Sensing.

[11]  A. Savitzky,et al.  Smoothing and Differentiation of Data by Simplified Least Squares Procedures. , 1964 .

[12]  A. J. Richardsons,et al.  DISTINGUISHING VEGETATION FROM SOIL BACKGROUND INFORMATION , 1977 .

[13]  M. Cho,et al.  A new technique for extracting the red edge position from hyperspectral data: The linear extrapolation method , 2006 .

[14]  S. Idso,et al.  Estimation of grain yields by remote sensing of crop senescence rates. , 1980 .

[15]  G. A. Blackburn,et al.  Spectral indices for estimating photosynthetic pigment concentrations: A test using senescent tree leaves , 1998 .

[16]  J. Dungan,et al.  The effect of a red leaf pigment on the relationship between red edge and chlorophyll concentration , 1991 .

[17]  Michael D. Steven,et al.  High resolution derivative spectra in remote sensing , 1990 .

[18]  Frédéric Baret,et al.  Modeled analysis of the biophysical nature of spectral shifts and comparison with information content of broad bands , 1992 .

[19]  G. Bonham-Carter.  Numerical procedures and computer program for fitting an inverted Gaussian model to vegetation reflectance data , 1988 .

[20]  Ruiliang Pu,et al.  Extraction of red edge optical parameters from Hyperion data for estimation of forest leaf area index , 2003, IEEE Trans. Geosci. Remote. Sens..

[21]  John R. Miller,et al.  Quantitative characterization of the vegetation red edge reflectance 1. An inverted-Gaussian reflectance model , 1990 .