Monitoring and Quantifying Canopy Growth of Paddy Rice (Oryza Sativa) using Ground-based Remote Sensing = 지상원격탐사를 이용한 벼의 군락생장 모니터링 및 정량화

다국어 초록 (Multilingual Abstract)

Monitoring growth conditions and phenological characteristics of crops could be achieved utilizing plant structural indices such as vegetation indices (VIs) as well as relationships between the scene reflectance values from remote sensing. Light emitting diodes (LEDs) are inexpensive and can be potentially used as a light transducer for remote sensing because it senses the same narrow band of wavelength which the LED emits. The objectives of this research were to develop an LED-based remote sensing system for monitoring crops and to determine the phenological stages of paddy rice and correlations of leaf area index (LAI) with VIs obtained from canopy reflectance. I determined phenological characteristics of paddy rice and correlations of LAI with VIs using canopy reflectance and LAI measured from a commercial spectra-radiometer and a leaf area meter, respectively, during the crop seasons of 2011-2013. I also fabricated an LED remote sensing system using Red LEDs (635 nm) and Near-Infrared LEDs (825 nm) to monitor canopy reflectance of crops. To monitor crop growth, normalized difference vegetation index (NDVI) was determined using reflectance data obtained from the LED remote sensing system and CropScan. The seasonal variation pattern of the LED remote sensing system NDVI was in good agreement with that of the CropScan NDVI. The root mean square difference (RMSD), mean relative difference (MRD), model efficiency (E) of the comparison between the NDVI values from the LED remote sensing system and CropScan were 0.07 %, 2.81 %, and 0.96, respectively. For NIR versus Red reflectance relationships, three phenological main stages (i.e., tillering, booting, and grain filling) were identifiable. During the tillering stage, Red reflectance decreased and NIR reflectance increased as number of tillers and chlorophyll pigments increased synchronously. Reflectance features in the grain filling stage showed an opposite trend to those in tillering. In the booting stage, Red reflectance remained relatively low (~ 3.52 %) and virtually invariant while NIR reflectance increased drastically (from 32.8 % to 47.8 %). Relationships between above ground dry weight (AGDW) and LAI values could be determined using an exponential function (R2 = 0.91). Relationships between VIs and LAI values could be also determined using exponential functions unanimously with reasonably acceptable ranges (0.83-0.92 for the VIs of interest) of R2 (coefficients of determination) values. These results allow to assume that it is possible to determine phenological aspects of rice growth qualitatively and quantitatively using remote sensing data.

목차 (Table of Contents)

• CHAPTER 1: Determining Canopy Growth Conditions of Paddy Rice based on Ground Remote Sensing 1
• 1.INTRODUCTION 1
• 2.MATERIALS AND METHODS 3
• 2.1 Paddy rice data 3
• 2.2 Remote sensing data 3
• 2.3 Statistical analysis and evaluation of regression models 4
• 3.RESULT AND DISCUSSION 5
• 3.1 Monitoring phenological stages 5
• 3.2 Monitoring canopy growth 6
• 3.2.1 Relationships of canopy growth or LAI with biomass 6
• 3.2.2 Relationships of LAI with VIs 6
• 4.CONCLUSIONS 8
• CHAPTER 2:Development of a Light Emitting Diode (LED)-based Remote Sensing System for Paddy Rice 20
• 1.INTRODUCTION 20
• 2.Description of the LED Remote Sensing System 21
• 2.1 Feature of LEDs 21
• 2.2 Structure of the LED remote sensing system 22
• 2.2.1 Up and down LEDs sensor head 22
• 2.2.2 Design of circuit 22
• 2.2.3 Data logger 23
• 3.Calibration of the LED Remote Sensing System 23
• 3.1 Modification of down-sensor position 23
• 3.2 Dark millivolt correction 24
• 3.3 Sun angle correction 24
• 4.Verification of LED Remote Sensing System Performance 25
• 5.Validation: Monitoring Growth of Paddy Rice 25
• 6.Summary and Conclusions 26
• REFERENCES 39
• ABSTRACT (in Korean) 44