Synergistic Inversion Of Rice Fpar Based On Optical And Radar Remote Sensing Data

Fraction of absorbed photosynthetically active radiation(FPAR)refers to the ratio of photosynthetically active radiation(PAR)absorbed by the green part of vegetation canopy to total PAR.FPAR is an important parameter that directly reflects the interception and absorption capacity of vegetation canopy for light energy.Accurate and quantitative acquisition of FPAR is of great significance for the study of terrestrial ecosystem processes,crop yield estimation,and carbon loss.For rice,FPAR is a key factor reflecting its growth status and health.The most common and efficient method for obtaining FPAR is using remote sensing data to estimate FPAR.Optical remote sensing data has been widely used in FPAR inversion.However,due to the cloudy and rainy climate of rice planting areas,optical remote sensing data is seriously insufficient.Radar remote sensing data has been used for inversion of rice biophysical parameters due to its all-time all-weather imaging capability.The revisit period of space-borne radar is usually long,for example,the RADARSAT-2 is 24 days.Therefore,it is difficult to monitor rice growth characteristics in time with only radar data.Field measurements were carried out in 2014 and 2016 in Meishan area of Sichuan Province.Optical and radar remote sensing data was acquired at the same time.Rice FPAR synergistic inversion model based on optical and radar data was constructed.The30-meter resolution rice FPAR product and the growth curve of rice FPAR was established.The research is mainly divided into the following steps:(1)Pre-processing the optical remote sensing data(GF-1,HJ-1A/B,Cbers-04,Landsat-8)to obtain reflectance data,and pre-processing RADARSAT-2 radar remote sensing data to obtain backscattering data of four polarization modes(VH,VV,HH,and HV).All data is filtered according to the data quality and data quantity to determine the optimal data for FPAR inversion.Through data analysis,it is found that the ratio of the two polarization modes VV/VH in the rice area changed a lot during the growth period,which can be used to extract the rice planting area.(2)based on the two-year ground measured FPAR data and remote sensing data,the statistical analysis is carried out to find the optimal vegetation index and radar backscattering coefficient combination,which is most suitable for FPAR inversion.The enhanced vegetation index EVI and radar backscattering coefficient combination(VH-VV)/(VV/VH)have good correlations with the measured FPAR.The R~2 for the EVI and the measured FPAR value is 0.8331.The R~2 for the(VH-VV)/(VV/VH)and the measured FPAR value is 0.7132.Fitting analysis were carried out between GF-1 and HJ-1A/B data,statistical models were established for several periods,based on which the remote sensing data of the whole rice growing season was unified.Using the principal component analysis method,the statistical model between FPAR and EVI,(VH-VV)/(VV/VH)is established for rice FPAR synergistic inversion with optical and radar data.FPAR estimated by the synergistic inversion model is obtained and the inversion accuracy is verified,the coefficient of determination R~2 reaches 0.8319.(3)Using the model of optical,radar and two data synergies to establish the FPAR curve of rice growth period,compared with the measured FPAR and MODIS FPAR.It is found that the results of the three models are better than the FPAR products of MODIS.The synergistic inversion model performed the best,the FPAR value is basically consistent with the measured data.According to the rice FPAR inversion model based on optical data,radar data and two data synergies,a 30-meter resolution rice FPAR product of rice growth period is established.