Estimating Rice Leaf Greenness (SPAD) Using Fixed-Point Continuous Observations of Visible Red and Near Infrared Narrow-Band Digital Images

Abstract

A narrow-band dual camera system demonstrated a new close-range sensing technique to seasonally track trends in leaf greenness in rice paddies. A weatherproof digital imaging system for the visible red (RED, 620—650 nm) and near infrared band (NIR, 820—900 nm) was positioned 12 m above a 600-m² rice field. During the 2009 and 2010 paddy rice seasons, the system automatically logged images at 10-min intervals throughout the day. Radiometric corrections for the images utilized solar irradiance sensors and prior calibration to calculate 0900-1500 JST daily-averaged reflectance factors (DARF). The DARF in RED (DARF-RED) and NIR (DARF-NIR) values were transformed to provide a daily-averaged normalized difference vegetation index (DA-NDVI). The DA-NDVI increased more rapidly in the vegetative growth period, and reached an asymptotic plateau earlier than the DARF-NIR. From transplanting to harvest, leaf greenness values (measured by the SPAD index) were measured for the central part of the uppermost leaves of targeted canopies weekly with a chlorophyll meter. We developed a leaf greenness index (LGI), the ratio of DA-NDVI to DARF-NIR, and a simple calculation method for area means to reduce the background effect. The modified area means of LGI followed the seasonal trend in SPAD value well; its pattern was inherently different from the patterns of any of the original three parameters: DARF-RED, DARF-NIR or DA-NDVI. Throughout the paddy seasons in the two years, a regression equation for estimating SPAD values using the LGI, daily solar radiation, the cosine of angle between the view and the meridian directions and the cosine of culmination solar zenith angle performed favorably (R²=0.815). The nitrogen concentration per dry plant hill (g kg^-1) had a close relation to the SPAD values estimated using the equation.