Estimating Of Rice Leaf Area Index From MODIS Data, Poyang Lake Area

The leaf area index(Leaf Areathe Index, abbreviationLAI)isoneoftheimportantparametersforcharacterization ofsurfacebiomass,vegetationfunction anditsecologicalstatuschange. Leafarea index of riceissensitivetohabitat changesin theenvironment, anddecides the yield of riceby influencingthephotosynthetic rate.PoyangLakePlain isa majorriceproduction bases. Toaccurateextractionleaf area indexoftherice inthis regionis of great significancefortheguidanceofagricultural productionandtherapid extractionofagricultural disasters.In this study,through the establishmentofavarietyofvegetation indexandsituleafarea indexby the regressionstatistics analysis,improved model ofestimating leaf area index. Andtheoptimal model isapplied“Estimating of Rice Leaf Area Index from MODIS Data, Poyang Lake Area",the results show thatthe rice leaf area index has differenttemporal and spatialvariationofdifferent growth stages.（1） Regression analysis between vegetation index and situ LAI showed that, inthe different growth stages of rice nonlinear estimation model is better than thelinear model. The best regre, in addition to the EVI is based on power estimationthe best rice reproductive growth period, the other vegetation indices arequadratic curve is the best, including the optimal estimation model for thequadratic model of GRNDVI.（2） At the different growth stages, the form of the best model of the wax ripestage is different from others’ whichbased on the logarithmic model is the best,the other growth stages and the reproductive growth consistent in order toestimate the effection of the quadratic curve best. Quite different but in a differentgrowing season, the best estimate model of the vegetation index, RVI a quadraticcurve to estimate the effect of the best in the panicle initiation stage, headingstage and ripening stage is of BNDVI a quadratic curve estimate best filling stageand milky stage EVI quadratic curve best mature stage in December, while EVILogarithim is the best.（3） The relative error analysis of the MODIS LAI and the optimal modelestimates LAI that this study, the optimal LAI estimation model accuracy than theaccuracy of the MODIS of LAI, the MODIS of LAI significant underestimationof the rice LAI exist. （4） Consistent with the ground the size of the quadrat, the high rate of theLandsat/ETM repair data on the spatial scale of MODIS push analysis found thatthe MODIS-NDVI and MOD09projected by the ETM extracted NDVI to LAIremote sensing estimation accuracy, show that the size of30m x30m quadratmeasured LAI extracted from MOD09vegetation index statistical regression isfeasible.（5） By usingthe optimal estimation model to estimate LAI in spatial abdicatedthatLAI near by Poyang Lake is smaller than where is far from the Lake.Although, the best of the different growth stages of rice LAI estimation modelaccuracy than the MODIS of LAI products of high, the statistical relationship of themodel itself is not very significant, the model estimates the error bit, and businessapplications for rice LAI estimation to be done to further validation study.