Parametric estimation of net photosynthesis in rice from in-situ spectral reflectance measurements

Net photosynthesis rate (PN) has an important role in controlling the ecosystem primary production and it is important to understand the relationships between PN and other physiological and environmental variables to improve and develop models for predicting plant growth and productivity. Here we study utility of ground-based hyperspectral measurements in estimating net photosynthesis during the growth cycle of rice. Ground-based hyperspectral reflectance and micrometeorological data (PN, chlorophyll concentration index (CCI), photosynthetically active radiation (PAR)) were collected from a rice field on three dates in the kharif season at Meerut, Uttar Pradesh, India. Green chlorophyll index (CIg) showed a positive and statistically significant relationship with CCI (R = 0.64). Use of single index showed poor fit with PN. The model based on a combination of two variables (vegetation index (VI) × PAR) however proved better in explaining the variance of PN. Among the VI-based combinations comprising three variables which depend on the Monteith light-use efficiency (LUE) concept, models consisting of either CIg or CCI as one variable appeared to be the best in predicting PN with R of nearly 0.6. The study concludes that a single index is insufficient to account for factors controlling variation in PN estimation whereas Monteith logic-based combinations have shown its potential for PN estimation by incorporating the chlorophyll based and other vegetation indices as the proxy of LUE and fAPAR (fraction of absorbed photosynthetically active radiation) respectively.