Quantitative Study On Estimation Of Crop Water Based On MODIS Data And Canopy Spectra Of Crop In North China

Drought has always had influences on yields of winter wheat in North China andmonitoring drought in North China has been paid more attentions by AgriculturalMeteorology and Remote Sensing researchers. Water plays an important role in crop life andit's also one of major factors to control photosynthesis and yields of crop. The remotedetermination of one of these biochemical constituents, crop water content, has importantimplications in agriculture, it is essential for drought assessment in natural vegetation.In this paper, characteristics of canopy spectrum and gorwth status of winter wheat underdifferent soil moisture levels were studied from the point of crop physiological drought usingexperimental data in the field. Correlations between FMC and EWT of canopy leaf andspectral reflectance of canopy were calculated and analysed quantitatively. On the basis ofquantitative analysis, the bands sensitive to canopy leaf water were found. Simple ratio waterindex and normalized difference water index was constructed with the sensitive bands. Thecorrelation coefficients between the index constructed in the study and FMC and EWT ofcanopy leaf were compared with the spectral indices founded in other studies. Simplestatistical models at different growth stages were established using spectral indices data andFMC and EWT of canopy leaf. A radiative transfer simulation, linking leaf and canopymodels was conducted and named PROSAIL model in this study. Canopy spectral reflectanceswere simulated with PROSAIL model and compared with the real canopy spectralreflectances in the field. The effect of leaf constituents and LAI on canopy canopy spectra inthe 400-2500nm region and spectral index NDWI was discussed. A modeling study isconducted, simulating canopy MODIS-equivalent synthetic spectra with random inputvariables and the statistical model that suited to the MODIS image was established. Canopyleaf water content EWT of winter wheat in North China at jointing and heading stages wasestimated using statistical model and MODIS data.The main outcomes in this study are as follows:(1) Due to the effect of chlorophyll content, water content, internal structure of leaf andcanopy structure etc, canopy reflectance of winter wheat has different sperturm repond atdifferent soil moistures levels and different growth stages.(2) Bands centered at 469,645,700 and 710nm of VIS region, bands centered at 760,815,855,930,1075,1100nm of NIR region and bands centered 550,1600,1640,1750,2130nm of SWIR were defined as sensitive bands to estimate leaf water content. These bandscenterd atmosphere windows had the potential to be applied in monitoring canopy leaf contentof crop. Simple ratio water index and normalized difference water index constructed with thesensitive bands could estimate canopy leaf content more accurately than single band. Theindices constructed in the study could estimate and monitor canopy water status of winterwheat.(3) The canopy reflectance of winter wheat was simulated with the mode. Results(R>0.98 and RMSE<0.05)showed grest correlation between the simulated and real canopyreflectance of winter wheat, demonstrating that radiative transfer model could potentially beused for simulating vegetation canopy reflectance, such as MODIS canopy reflectance. Theapplicability of radiative transfer model inversion methods to MODIS is studied, investigatingits spectral capability for water content estimation.