Estimation Model Based On Hyperspectral Remote Sensing For Biophysical And Biochemical Parameters Of Amphor

We mainly studied reflectance spectra characteristics and biophysical and biochemical parameters of camphor based on the field and canopy spectral data, measured by EPP HandHeld Spectrometer.The main contents and conclusions are summarized as follows:(1). Spectral discrepancy was found in spectral green peaks, red valley and near infrared area from camphor in different months. There was not obvious differences in other bands from reflectance spectrum. derivative spectrum can clearly reflect the variation characteristics camphor tree spectrum; The phenomenon from pectrum curve displacement is not obvious.(2). Spectral discrepancy was found in spectral green peaks, red valley and near infrared area from camphor in different SPAD and water contents. Reflectance spectrum decreased along with the increasing SPAD, Reflectance spectrum increased along with leaf water contents increasing.(3). We analyzed biophysical and biochemical information using remote sening based on reflectance spectrum and derivative spectrum from camphor canopy.The results indicated:derivative spectrum has better correlated than reflectance spectrum, although showed great volatility, but derivative spectrum showed greater regular with biochemical parameters:400nm-550nm, showed regularly increase or decrease,550nm-680nm appeared the opposite rule,750nm later, showed great fluctuation.(4). LAI,CHLa,CHLb,CHLa+b,N,P showed better correlated with single spectrum parameters, and the largest correlation was on single spectrum parameters. However, CAR,EWT showed better correlated with composite spectrum parameters, and composite spectrum parameters was better than single spectrum parameters in correlation. Ratio vegetation index and the normalized vegetation index had better correlated with biophysical and biochemical parameters, they will enlarge the differences of biochemical variables.(5). CHLa,CHLa+b,CHLb,N,P showed the maximum correlation coefficient at Dr, CAR,EWT expressed their maximum correlation coefficient including SDr, Dr is the maximum in red edge of derivative spectrum within. SDr is value combined with red edge wavelength range derivative spectrum. this is the reason that many scholars choose red edge parameters.(6). Many vegetation index were put forward for the remote sensors, hyperspectral remote sensing had many information about data, data intensive, hyperspectral data showed not sensitive for the traditional definition of remote sensing spectrum.It was the time to develop the vegetation index for hyperspectral data.(7). The suitable models are following:CHL-a+b content:selected band R750/R708,Y=2.828x2-5.188x+3.258, R2=0.397;EWT content:selected band (SDr-SDb)/(SDr+SDb), Y=24.77x2+16.54x+9.478, R2=0.562;LAI content:selected band Ro, Y=722.2x2+10.51x+4.368, R2=0.493;N content:selected band Dr, Y=-32543x2-235.9x+24.48, R2=0.50;P content:selected band Rg, Y=-0.18In(x)+0.155, R2=0.522。...