Monitoring of Rice Growth Using Polarimetric Scatterometer System RSSJ 2009, Komaba Research Campus, May 21-22, 2009 Monitoring of Rice Growth Using Polarimetric Scatterometer System Yihyun Kim1*, S.Young Hong1, Eunyoung Choe1 and Hoonyol Lee2 1 National Academy of Agricultural Science, RDA, Korea 2 Department of Geophysics, Kangwon National University, Korea Good afternoon My name is Yi-Hyun Kim from National Institute of Agricultural Science Technology in Korea My presentation title is radar backscattering measurement of a paddy rice field using multi-frequency(L, C and X) and full-polarization  

Background Rice is one of the major crops in Korea Microwave radar can penetrate cloud cover regardless of weather condition Ground-based polarimetric scatterometer has advantage of monitoring crop conditions with full polarization and various frequencies Plant parameters such as LAI, biomass, plant height are highly correlated with backscattering coefficients ENVISAT SAR data (5.3 GHz, hh-, hv-polarizations, and incidence angles between 28.5° and 40.9°) to monitor rice growth and compared the data with simulation results (Le Toan et al, 1997) RADARSAT data (5.3 GHz, hh-polarization, and incidence angles between 36° and 46°) was analyzed for monitoring the rice growth in Korea (Hong et al, 2000) Rice is one of the major crops in Korea Microwave radar sensing has great potential, especially in monsoon Asia, since optical observations are often hampered by cloudy conditions Especially a ground-based polarimetric scatterometer has advantage of monitoring crop conditions continuously using full polarization and various frequencies Many plant parameters such as LAI, biomass, plant height are highly correlated with backscattering coefficients. According to frequency, polarization between plant parameters and backscattering coefficients was different For example last study work, Le Toan was study ENVISAT SAR data (5.3 GHz, hh-, hv-polarizations, and incidence angles between 28.5° and 40.9°) to monitor rice growth and compared the data with simulation results and Hong was study RADARSAT data (5.3 GHz, hh-polarization, and incidence angles between 36° and 46°) was analyzed for monitoring the rice growth in Korea

Objective To analyze scattering characteristics of paddy rice obtained from polarimetric scatterometer system in 2007-2008 Relationship between backscattering coefficients in L, C, X-band and rice growth variables with full polarization and various angles We measure backscattering coefficients of paddy rice using L, C, X-bands scatterometer system during the growth period and investigate relationship between backscattering coefficients and rice growth variables for example plant height, Leaf Area Index, Biomass with full polarization and various angles

Radar Backscattering Measurement of a Paddy Rice Field using Multi-frequency (L, C and X) and Full-polarization (2007) Study site Construction of polarimetric scatterometer system (L, C, X-band) Calculation of backscattering coefficients Measurement of backscattering coefficient for L, C, X-band during rice growth stage In conclusion Backscattering coefficients of rice crop were measured with a ground-based scatterometer. The measurement was carried out at L-, C-, and X-band with full polarizations and different incident angles. The temporal variations of the backscattering coefficients of the rice crop at L-, C-, and X-band during a rice growth period. At larger incident angles, range of backscattering coefficients were higher than that of small incident angle. VV polarization backscattering coefficients were higher than HH-polarized backscattering coefficients in early rice growth stage. We conducted the relationship between backscattering coefficients with L-, C-, and X-band and rice growth parameters. Biomass was correlated with L-band hh-polarization at a larger incident angle. LAI was highly correlated with C band hh- and cross-polarizations. Grain weight was correlated with backscattering coefficients with X-band VV polarization at a larger incidence angle. X-band was sensitive to grain maturity at near harvesting season.

Study site The size field : 660m2 Location : Experimental field at NAAS, Suwon, Korea Testing varieties : Chuchoungbyeo The size field : 660m2 The test site was located at a NAAS experimental field Suwon, Korea. The rice cultivar was a Japonica type, called Chuchoung. The size field was about 660m2

Dual Polarimetric Horn Antenna Polarimetric scatterometer system Consist of polarimetric scatterometer system Dual Polarimetric Horn Antenna Network Analyzer Polarimetric scatterometer system consists of dual polarimetric horn antenna, vector network analyzer (VNA), RF cables, and a personal computer The system is calibrated using a calibration kit The VNA-based polarimetric scatterometer operates in a stepped-frequency sweep mode Polarimetric scatterometer provides a time domain radar return from a target as a fully polarimetric amplitude and phase data L-band C-band X-band Calibration kit

Dual polarimetric horn Specification of the scatterometer system Specification L-Band C-Band X-Band Center frequency 1.27GHz 5.3GHz 9.65GHz Bandwidth 0.12GHz 0.6GHz 1GHz Number of frequency points 201 801 1601 Antenna type Dual polarimetric horn Antenna gain 12.4dB 20.1dB 22.4dB Polarization HH, VV, HV, VH Incident angle 20° ~ 60° Platform height 4.16m This system equipped with L, C, X-bands, and with full polarization The incident angles were between 20~60

Calculation of backscattering coefficients (apply to radar equation) Backscattering coefficients were calculated by applying radar equation. It defined as the following expression. Backscattering coefficients of each band were calculated as the follow expressions. Backscattering coefficients of each band the follow expression

Temporal variations of backscattering coefficients in bands Temporal variations of backscattering coefficients at polarization and incident angle 30°~60° for the L-band Incident angle: 30 Incident angle: 40 Heading Radar backscattering measurements were about 5 months from before transplantation of the rice seedings to harvest of the mature rice crop. This figure shows the temporal variations of the backscattering coefficients of the rice crop at L-band after transplanting, at various incidence angles. Backscattering coefficients of paddy fields at L-band ranged from were about -55dB ~ 0dB. VV-polarized backscattering coefficients higher than hh- and hv/vh-polarized backscattering coefficients during rooting stage.  Cross polarized backscattering coefficients increased towards the heading stage (mid-Aug) and thereafter saturated, again increased near the harvesting season. Incident angle: 50 Incident angle: 60

Temporal variations of backscattering coefficients in bands Temporal variations of backscattering coefficients at polarization and incident angle 30°~60° for the C-band Incident angle: 40 Incident angle: 30 heading Incident angle: 50 This figure shows changes of backscattering coefficients at C-band during the growing period. The HH-polarized backscattering coefficients at all incident angles (except 20°) increased as growth advanced and saturate at the ripening stage. Incident angle: 60

Temporal variations of backscattering coefficients in bands Temporal variations of backscattering coefficients at polarization and incident angle 30°~60° for the X-band Incident angle: 30 Incident angle: 40 heading This figure shows change in backscattering coefficients at X-band with growth   Backscattering coefficients of range at X-band lower than that of L-, C-band. HH-, VV-polarized σ° steadily increased toward panicle initiation stage and thereafter decreased, and again increased near the harvesting season. This dual-peak trend was clearer larger incident angles. Larger incident angles (over 50°) at cross-polarized σ° showed similar phenomena.   Fresh weight was decreased and heads of the canopy were easily show, so X-band as high frequency sensitive to heading or grain maturity after heading stage. Incident angle: 50 Incident angle: 60

Analysis of scattering characteristics in paddy rice by X-band automatic scatterometer system (2008) Construction of X-band automatic scatterometer system Growth data collection Measurement of backscattering coefficient for X-band during rice growth stage In conclusion Backscattering coefficients of rice crop were measured with a ground-based scatterometer. The measurement was carried out at L-, C-, and X-band with full polarizations and different incident angles. The temporal variations of the backscattering coefficients of the rice crop at L-, C-, and X-band during a rice growth period. At larger incident angles, range of backscattering coefficients were higher than that of small incident angle. VV polarization backscattering coefficients were higher than HH-polarized backscattering coefficients in early rice growth stage. We conducted the relationship between backscattering coefficients with L-, C-, and X-band and rice growth parameters. Biomass was correlated with L-band hh-polarization at a larger incident angle. LAI was highly correlated with C band hh- and cross-polarizations. Grain weight was correlated with backscattering coefficients with X-band VV polarization at a larger incidence angle. X-band was sensitive to grain maturity at near harvesting season.

X-band automatic scatterometer system Construction of X-band automatic scatterometer system Weather condition (such as precipitation, wind, humidity) negative effect

Specification of the automatic scatterometer system X-Band Center frequency 9.65GHz Bandwidth 1GHz Number of frequency points 1601 Wavelength 0.031m Slant range resolution 0.15m Antenna type Dual polarimetric horn Antenna gain 22.4dB Polarization HH, VV, HV, VH Incident angle 45° Measurement interval 1 per 10minutes

Growth data collection Plant height, Fresh and dry weight, LAI Collection interval : about 1 per a week Transplant stage (mid-May) Panicle formation stage (mid-July) Heading stage (mid-Aug) Harvesting stage (mid-Oct) Growth data for the rice canopy, such as LAI, fresh and dry weight and plant height, were acquired at time of each scatterometer measurement simultaneously

Temporal variations of backscattering coefficients in X-band Temporal variations of backscattering coefficients at polarization and incident angle 45° for the X-band Incident angle: 45 2007 2008

Relationship between backscattering coefficients at L, C, X-band and rice growth parameters (2007-2008) Relationship between backscattering in bands and plant parameters during rice growth stage Optimum condition between backscattering coefficients in bands and plant parameters

Relationship between backscattering coefficients in bands and rice growth parameters Correlation between backscattering coefficients at L-band and plant variables VV HH HV Incident angle Plant height LAI Tfw (g/m2) Tdw 20 -0.93*** -0.81** -0.90*** -0.87*** -0.56* -0.20ns -0.37ns -0.32ns 0.21ns 0.38ns 0.29ns 0.32ns 25 -0.53* 0.24ns 0.44* 0.76** 0.85** 0.81** 30 -0.01ns 0.28ns 0.15ns 0.18ns -0.39ns -0.38ns -0.42* -0.40ns 0.91*** 0.77** 0.82** 35 -0.49* -0.63* -0.58* -0.62* 0.40ns 0.25ns 0.31ns 0.89*** 0.71** 0.80** 0.78** 40 0.58* 0.70* 0.68* 0.74** 0.73** 0.72** 0.86*** 45 0.92*** 0.87*** 0.94*** 0.90*** 50 0.63* 0.75** 0.97*** 0.98*** 55 0.62* 0.93*** 0.88*** 60 0.79** 0.84** * : level of significance p<0.05 ** : level of significance p<0.01 *** : level of significance p<0.001 Tfw : Total fresh weight, Tdw : Total dry weight

Relationship between backscattering coefficients in bands and rice growth parameters Correlation between backscattering coefficients at C-band and plant variables VV HH HV Incident angle Plant height LAI Tfw (g/m2) Tdw 20 -0.94*** -0.74*** -0.84** -0.83** -0.67* -0.76** -0.75** 0.81** 0.67* 0.74** 0.71** 25 0.72** 0.75** 0.48* 0.46* 0.85** 0.92*** 0.90*** 30 0.82** 0.78** 0.86*** 0.83** 0.84*** 0.91*** 35 0.70* 0.68* 0.93*** 0.84** 0.89*** 0.88*** 40 0.38ns 0.55* 0.50* 0.95*** 0.87*** 45 0.64* 0.56* 0.58* 50 0.76** 0.73** 0.94*** 55 60 0.44* 0.43*

Relationship between backscattering coefficients in bands and rice growth parameters Correlation between backscattering coefficients at X-band and plant variables VV HH HV Incident angle Plant height LAI Tfw (g/m2) Tdw 20 0.26ns 0.41* 0.32ns 0.68* 0.63* 0.64* 0.80** 0.83** 0.82** 25 0.62* 0.70* 0.67* 0.72** 0.66* 0.73** 0.74** 30 0.46* 0.57* 0.54* 0.52* 0.84** 0.65* 0.75** 0.70** 0.69** 35 0.50* 0.61* 0.81** 0.86*** 0.85** 0.71** 0.79** 40 0.43* 0.55* 0.56* 45 0.33ns 0.45* 0.42* 0.40* 0.76** 50 0.23ns 0.29ns 0.28ns 0.24ns 0.77* 55 -0.10ns -0.20ns -0.13ns -0.18ns 60 -0.25ns -0.44* -0.36ns -0.41* 0.78** 0.77**

Relationship between backscattering coefficients in bands and rice growth parameters Correlation between backscattering coefficients at bands and grain dry weight L-band(2007) C-band(2007) X-band(2007) X-band(2008) Incident angle VV HH HV 20 -0.96*** -0.85** -0.64* -0.50* -0.19ns 0.26ns -0.54* -0.05ns 0.10ns 25 -0.97*** -0.74** 0.06ns -0.39ns -0.70* -0.33ns 0.35ns 0.56* 30 -0.78** 0.53* -0.55* -0.38ns 0.51* 0.31ns -0.45* 35 0.43* 0.72* -0.81** -0.32ns 0.27ns 0.70** -0.30ns -0.40* 40 0.61* 0.40* 0.66* -0.22ns -0.86*** 0.78* -0.36ns 45 0.75** 0.23ns 0.63* -0.83** -0.13ns 0.88*** 0.39ns 0.58* 0.89*** 0.21ns 0.44* 50 0.71* -0.67* 0.29ns -0.16ns -0.52* 0.84** 0.55* 0.65* 55 -0.29ns 0.07ns 0.17ns -0.77** 0.80** 0.70* 60 0.30ns 0.18ns -0.10ns 0.67* 0.81** 0.74** 0.69*

Coefficient of determination Optimum condition between backscattering coefficients in bands and plant parametrs Optimum condition Plant parameters Band Polarization Incident angle Coefficient of determination Plant height(cm) C-band HV 45° R2=0.90** LAI HH 50° R2=0.91** Biomass(g/m2) L-band R2=0.93** Grain dry weight(g/m2) X-band VV R2=0.82**

Conclusions Backscattering coefficients of rice crop were measured with ground-based scatterometer The temporal variations of the backscattering coefficients of the rice crop at L-, C-, X-band with full polarization during rice growth period VV-polarization backscattering coefficients higher than HH-polarization backscattering coefficients in early rice growth stage Relationships between backscattering coefficients and the plant parameters Biomass was correlated with L-band HH-polarization at large incident angle LAI was highly correlated with the C-band HH- and cross-polarizations X-band was sensitive to grain maturity at near harvesting stage Optimum condition between backscattering coefficients in bands and plant parameters In conclusion Backscattering coefficients of rice crop were measured with a ground-based scatterometer. The measurement was carried out at L-, C-, and X-band with full polarizations and different incident angles. The temporal variations of the backscattering coefficients of the rice crop at L-, C-, and X-band during a rice growth period. At larger incident angles, range of backscattering coefficients were higher than that of small incident angle. VV polarization backscattering coefficients were higher than HH-polarized backscattering coefficients in early rice growth stage. We conducted the relationship between backscattering coefficients with L-, C-, and X-band and rice growth parameters. Biomass was correlated with L-band hh-polarization at a larger incident angle. LAI was highly correlated with C band hh- and cross-polarizations. Grain weight was correlated with backscattering coefficients with X-band VV polarization at a larger incidence angle. X-band was sensitive to grain maturity at near harvesting season.

Thank you RURAL DEVELOPMENT ADMINISTRATION (RDA)