Crop damage assessment after toluene vapor exposure on Oryza Sativa L. based on Vegetation Indices (VIs) extracted from Unmanned Aerial Vehicle (UAV) multispectral imagery

Abstract

The annual number of chemical accidents is increasing as the amount of chemical used in industry in South Korea increases. The chemical accident causes damage to plants by exposure to spilled hazardous chemicals. However, the conventional methods for crop damage assessment after the chemical accident have limitations. Visual inspection and sampling can be inaccurate, time-consuming, and subjective. As well as, modeling can be inaccurate according to the initial parameters. This study focused on the application of Vegetation Index (VI) extracted from Unmanned Aerial Vehicle (UAV) multispectral imagery for crop damage assessment after the chemical accident. The objective of the study was to evaluate the feasibility of the usage of VI extracted from UAV multispectral imagery for crop damage assessment after chemical exposure. This study simulated the chemical accident by exposure of toluene to Oryza Sativa L. (Asian rice) at five growth stages with four levels of toluene concentrations. A multispectral camera mounted on UAV was used to capture the multispectral imagery. For damage assessment after the toluene exposure, the flight missions were conducted five days after toluene exposure (Day After Damaged, DAD) at each growth stage. The toluene exposed rice at each growth stage was re-transplanted to the paddy field to investigate recovery after toluene exposure. For recovery assessment, the flight mission was conducted 67 days after planted (DAP). Normalized Difference Vegetation Index (NDVI), Green Normalized Difference Vegetation Index (GNDVI), and SAVI (Soil-adjusted Vegetation Index) of rice with different toluene treatment were computed based on UAV multispectral imagery. Physiological characteristics: leaf chlorophyll content and grain yield were also measured. Among the three VIs, NDVI showed the highest sensitivity with respect to the toluene exposure on rice. The mean NDVI value of toluene exposed rice measured at 5 DAD was significantly decreased according to the toluene concentration increase except for the toluene exposed rice at the flowering stage. The mean NDVI value of re-transplanted rice measured at 67 DAP was lower at the rice exposed to the higher concentration of toluene at the earlier growth stage. The leaf chlorophyll content of toluene exposed rice was significantly decreased according to the toluene concentration increase except for the toluene exposed rice at the booting and flowering stages. The grain yield of rice was lower at the toluene exposed rice exposed at all growth stages. The NDVI showed positive correlation with leaf chlorophyll content (R=0.550, p=0.0046 for leaf chlorophyll content, R=0.652, p=0.0007 for SPAD value) and grain yield (R=0.645, p=0.0127). The correlation analysis results indicate the strong relationship between remotely-sensed data (NDVI) and ground-based data (leaf chlorophyll content, grain yield).