A framework for rice heavy metal stress monitoring based on phenological phase space and temporal profile analysis

Abstract

Previous studies make it possible to use remote sensing techniques to monitor heavy metal stress of rice synchronously and continuously. However, most studies mainly focus on the analysis of rice’s visual symptoms and physiological functions rather than temporal information during the growth period, which may reflect significant changes of rice under heavy metal stress. In this paper, an enhanced spatial and temporal adaptive reflectance fusion model was used to generate synthetic Landsat time series. A normalized difference water index and an enhanced vegetation index were employed to build phenological phase space. Then, the ratio of the rice growth rate fluctuation (GRFI Ratio) was constructed for discriminating the different heavy metal stress levels on rice. Results suggested that the trajectories of rice growth in phenological phase space can depict the similarities and differences of rice growth under different heavy metal stress levels. The most common phenological parameters in the phase space cannot accurately discriminate the heavy metal stress level. However, the GRFI Ratio that we proposed outperformed in discriminating different levels of heavy metal stress. This study suggests that this framework of detecting the heavy metal pollution in paddy filed based on phenological phase space and temporal profile analysis is promising.