Monitoring Rice Phenology And Growth Parameters Using Near-Ground Remote Sensing Platforms

Remote sensing(RS)technique could be used to acquire crop growth status information rapidly,nondestructively and real-timely,which provides reliable technical support for precision agriculture.Near-ground based RS platforms are easily operable and flexible,which could offer variable data for crop biophisical and biochemical parameters estimation and improve the accuracy and reliability for crop growth status monitoring.In this study,four rice field experiments were conducted,involving different growing seasons,rice cultivars,planting densities and nitrogen rates.Two portable spectrometers(ASD FieldSpec Pro spectrometer and GreenSeeker)were used to collect rice canopy reflectance and variable sensors(RGB,color infrared(CIR)and multispectral(MS)cameras)mounted on unmanned aerial vehicle(UAV)were applied to obtain rice canopy imagery.We explored the rapid detection method for rice key phenological dates with time series of spectral indices(Normalized difference vegetation index,NDVI and red edge chlorophyll index,CIred edge),investigated the performance of different UAV sensors(RGB,CIR and MS cameras)on rice leaf and plant nitrogen accumualtion(LNA and PNA)estimation,determined the optimal spectral and textural index in rice biomass and plant N concentration(PNC)estimation,established monitoring models for aboveground biomass(AGB)with the combination of spectral index and textural index both from MS images,and monitoring models for PNC with the combination of ground-based spectral index and textural index from MS images.The results would be helpful to provide theoretical foundation and technical support for crop growth status monitoring with near ground remote sensing.Firstly,a phenological detection method was determined based on time series analysis of ground-based spectral index data.A comparison of remote sensing-based estimates with field observations over three years with different cultivars,planting densities and nitrogen(N)rates showed that NDVI from both spectrometers can be used to detect the dates of critical tillering,middle heading and maturity.Specifically,NDVIGS yielded better performance than NDVIASD for estimating the three phenological dates.CIred edge can accurately estimate the dates of jointing,middle booting and dough grain.This work has great potential to provide valuable support for assessing crop growth status and providing precise management strategy.The dates of critical tillering,jointing and maturity detected from a combination of CIred edge and NDVI could be useful for irrigation and fertilization management,and harvest determination,respectively.Secondly,the quantitative relationships between spectral indices extracted from different UAV imageries and rice LNA and PNA were analyzed,and the application of different UAV sensors on LNA and PNA estimation was determined.Results demonstrated that the red edge indices derived from MS images produced the highest estimation accuracy for LNA(R2:0.79-0.81,RMSE:1.43-1.45 g m-2)and PNA(R2:0.81-0.84,RMSE:2.27-2.38 g m-2).GNDVI from CIR images yielded a moderate estimation accuracy with an all-stage model.Color indices from RGB images exhibited satisfactory performance for the combination dataset of the tillering and jointing stages.Compared with the counterpart indices from digital images,the indices from MS images performed better in most cases.Then rice AGB estimation model was established with the combination of spectral and textural indices,based on the analysis of quantitative relationships between rice AGB and spectral and textural indices extracted from UAV MS imagery.Results indicated that optimized soil adjusted vegetation index(OSAVI)exhibited the best relationship with AGB for the whole season(R2=0.63)and post-heading stages(R2=0.65),and red-edge indices yielded best performance(R2>0.70)only for the growth stages before heading.The texture measurement mean in NIR(MEA800)band was the optimal texture among the eight texture measurements in AGB estimation.Texture index(NDTI(MEAsoo,MEA550))was superior to all the evaluated ?s in estimating AGB for the whole season(R2=0.75)and pre-heading stages(R2=0.84).Further improvement was obtained across the whole season by combining NDTI(MEA800,MEA550),GNDVI and MTVI2 through a multiple linear regression.This multivariate model produced the highest estimation accuracy for all stages(R2=0.83 and RMSE=1.68 t ha-1)and different stage groups(R2=0.87 and RMSE=0.99 t ha-1 for pre-heading stages and R2=0.75 and RMSE=1.73 t ha-1 for post-heading stages).Finally,after comparing the performance of ground hyperpsectral data and UAV multispectral imagery on rice PNC estimation,we found that aerial ?s performed well only for pre-heading stages(R2=0.52-0.70),and photochemical reflectance index and blue N index from ground(PRIg and BNIg)performed consistently well across all growth stages(R2=0.48-0.65&0.39-0.68).Most texture measurements were weakly related to PNC,but the optimal NDTIs could explain 61%and 51%variability of PNC for separated stage groups and entire season,respectively.Moreover,multiple linear regression models combining aerial?s and NDTIs did not significantly improve the accuracy of PNC estimation,while models composed of BNIg and optimal NDTIs exhibited significant improvement for PNC estimation across all growth stages.Therefore,the integration of ground-based narrow band spectral indices with UAV-based textural information might be a promising technique in crop growth monitoring.