Model Study On Winter Wheat Vegetation Reflection Character And Leaf Area Index Based On High Spectrum Remote Sensing

The precision agriculture is important way for agriculture to realize the low consumption, highly effective, high quality, the environmental protection. So, at present precision agriculture is the new tendency for the world agriculture development, and also is optimal path to the 21st century for our country agriculture.The precision agriculture is a new agriculture comprising the present information technology, the biological technology and the engineering technology, which is based on a series of high technology and new technology newest achievement. The research on the precision agriculture has been the key of the world agriculture technology, of which the remote sensing technique is one of the important tools to realize accurate agriculture.This paper analyzed the winter wheat physiology ecology index under different water treatment and density of sowing by the method of the experiment and the theoretical analysis, and obtained the winter wheat Spectra reflectivity under different development period using the spectroscope to establish leaf area index based on high spectrum remote sensing. The main conclusions are as follows:(1)The spectrum curve characteristic (in visible light scope) has close relation with the chlorophyll comparative content, when the chlorophyll comparative content is higher, the corresponding spectrum curve index of reflection will be smaller; When the chlorophyll comparative content is lower, the corresponding spectrum index of reflection is higher.The chlorophyll content is higher, the red side oscillation amplitude is bigger, the red side area is also bigger, the red side position is more close to the long wave direction Along with the period of duration advancement, the red side moves to the long wave then moves to short wave, the area of red side increases first and then decreases, the red side oscillation shows the trend with increases first and then decreases(2)There exists between good negative correlation relations between spectrum index of reflection and winter wheat's leaf area index in 350-690nm; There exists between good positive correlation relations between spectrum index of reflection and winter wheat's leaf area index in 750-960nm. Analyzing correlation relations between 10 common vegetation indices and the winter wheat leaf area index, the result showed that correlation relations between normalization vegetation index and the winter wheat leaf area index relevance is best, and R~2 of the model is highest, its model is y=3.1437x2.2152, R2=0.8160; the next is that correlation relations between ratio vegetation index and the winter wheat leaf area index relevance is best is also quite good, its model form is y=0.9078Ln (x) - 0.1734, R2=0.80601(3)Introduced the neural network model to estimate the winter wheat's leaf area index, and contrasted forecast precision of two kind of neural network model, the relative error absolute value is 1%-18% using BP neural network model, the model forecast's average error was 11.55%, the relative error absolute value is 11.55%, the model forecast's average error was 7.15% using RBF neural network model, , obviously precision of the RBF neural network is higher than that of the BP neural network.(4)The density of sowing and the water treatment have the obvious influence to winter wheat's high and the chlorophyll content. Under the same density of sowing condition, the influence of different water treatments to winter wheat's high and the chlorophyll is that winter wheat's high growth rate and the chlorophyll comparative content under the suitable water supply condition are higher than that under mildly water loss condition , and the latter is bigger than that under severe water loss condition.Under the same water treatment, the influence of different water treatments to winter wheat's high and the chlorophyll is that winter wheat's high growth rate and the chlorophyll comparative content under the light density of sowing condition are higher than that under the suitable density of sowing condition , and the latter is bigger than that under the heavy density of sowing condition.(5)The winter wheat leaf area index showed the tendency of rising first and falling later no matter what treatment condition, under the same density of sowing condition, with the water deficit increase of leaf area index smaller.(6)Established the winter wheat leaf area index change model under different treatnents, of which, the predictable average error is 12.12% by the multi-dimensional linear regression, the predictable average error is 5% by gray GM (1,1) , the predictable average error is 5.55% by the BP neural network, obviously, the predition of leaf area index using the gray forecast model and the BP neural network is able to obtain the good result.(7)Analysis result of output under different treatments showed that the density of sowing plays a crucial role in the output of wheat, while output is proportionate to irrigation amount under the same density of sowing .