Estimating paddy rice leaf area index with fixed point continuous observation of near infrared refl ectance using a calibrated digital camera

Abstract

A two-band digital imaging system-one band for visible red (RED, 630-670 nm) and the other for near infrared (NIR, 820-900 nm)-was positioned 12 m above a 600-m2 zice fi eld. The imaging system automatically logged bird's-eye-view images at 10-min intervals from 0700 to 1700 JST daily during the 2008 paddy rice season. Radiometric corrections and midterm field validations for the pairs of two-band images utilized solar irradiance sensors and prior calibrations to calculate 0900-1500 JST daily-averaged refl ectance factor (DARF). The DARF values in the RED and the NIR bands agreed with observations made with a portable spectroradiometer and showed smooth seasonal changes in terms of plant growth. During the before-heading period, NIR DARF values had a correlation coefficient (r) of 0.91 with the leaf area index (LAI). An empirically derived equation for LAI using NIR DARF values and the cosines of angles between the view and the planting row directions, and the view and the meridian directions, showed R2 values of 0.93, in estimations of LAI from the dataset (number of observations=52) acquired at four sampe points centrally located within the viewing field. Validation indicated that the equation also worked well for the other six observation points spread across the viewing field, the data of which were not used in deriving the equation parameters. The DARF values observed in the NIR band calibrated with solar spectral irradiance sensors were useful for assessing rice LAI during the before-heading periods.