| Nitrogen application is one of the most important links in rice production,but there are some problems in rice production in China,such as excessive nitrogenous fertilizer and low utilization rate.Excessive nitrogenous fertilizer not only increases the production cost,but also pollutes the environment.Therefore,it is of great significance to study the precision management of nitrogen fertilizer in rice production.The rapid acquisition and analysis of rice growth information and the corresponding fertilization decision model are the key to realize the precise management of nitrogen fertilizer in rice.The UAV remote sensing has the characteristics of low operating cost,flexible operation,low reliance on landing sites and large sampling time scale,which can make up the defects of existing aerospace and aeronautical remote sensing technology,and is widely used in the rapid acquisition of crop growth information.But at present,there are many factors affecting rice nitrogen uptake and utilization,and the difference of growth information between different varieties of rice in different growing period.To address the above problems,this subject takes the Meixiangzhan II rice variety as the test object,uses the ground remote sensing(GreenSeeker hand-held spectrometer,the chlorophyll meter SPAD)and the UAV Remote Sensing(ADC Lite multispectral camera)data mutually verifies the method,unifies this variety the characteristic and the matching agronomy,the local planting habit and so on.In this paper,the study of rice ground remote sensing and UAV Remote sensing information analysis,rice nitrogen nutrition diagnosis and decision making model was carried out in field experiment,and a supporting system for accurate management of nitrogen nutrition in rice was developed.(1)Based on the analysis of the ground remote sensing information of rice under different nitrogen application levels,and using the "3414" fertilizer test method,4 rice planting experiments with different nitrogen gradients were designed,and the SPAD502 and GreenSeeker RT 100 instruments were used to collect data at approximately 6d intervals,The results of data analysis showed that the NDVI of rice canopy was significantly correlated with the SPAD value of chlorophyll content in rice leaves.The correlation coefficients of each period reached more than 0.6,of which the coefficient of correlation between 0.836;NDVI and effective accumulated temperature was up to 0.996 in tillering period and spike and milk maturity stage,and the determining coefficients of tillering period(0-30 d after transplanting days),spike and milk maturity stage(42-87 d after transplanting days)were respectively up to 0.996 and 0.953,showed that the growth of rice and effective accumulated temperature is closely related to the establishment of standard nitrogen fertilization in the effective accumulation temperature and the rice canopy NDVI analytic model,is advantageous in the rice application nitrogen level diagnosis,avoids because of the geographical difference causes the climate influence factor to the model interference.(2)In order to solve the problem of the dynamic change of canopy NDVI in different growing period,it is difficult to establish a stable analytic model,and a method for the determination of nitrogen level in rice with RISP(Ratio index of standard planting)is proposed.This method compared the NDVI of rice in the season with that of standard nitrogen fertilization,and realized the diagnosis of nitrogen application.DISP(Differencevalue index of standard planting)and RISP were constructed,Selecting the coefficient of variation and regression model as the evaluation index to compare the two indexes,the results show that: in the whole growth period of rice rice,the RISP is more stable than DISP coefficient of variation and high determining coefficient,which indicates that RISP is more steady;based on RISP,the diagnostic model of nitrogen level in rice was established.The determining coefficient of the regression model is 0.991,and the diagnostic model of nitrogen fertilization can be established without considering the factors such as soil physical and chemical properties.(3)The experiment of nitrogen nutrition diagnosis between the UAV remote sensing and ground remote sensing was carried out,and the decision model of rice variable nitrogen application based on UAV remote sensing was established.In the top dressing period,the fixed-wing UAV remote sensing platform(aerial height 200m)and the ground remote sensing greenseeker RT 100 equipment(from canopy 80cm)were used to collect the NDVI data of rice canopy,and the correlation between them was studied,The results show that: the correlation between the remote sensing NDVI and the terrestrial remote sensing NDVI is very significant(at 0.01 significant level)and the correlation coefficient is 0.897,which indicates that the rice nitrogen nutrition diagnosis model based on the ground remote sensing NDVI can be directly used in the remote sensing rice nitrogen nutrition diagnosis model.Based on the nitrogen nutrition diagnosis model of rice based on the UAV remote sensing NDVI,a rice nitrogen nutrition diagnosis model and its variable nitrogen application decision model were established.(4)Using the established decision model of remote sensing rice variable nitrogen fertilization,plot tests and field tests were arranged in Guangzhou Cencun Farm,Luoding Taiping Town and Nanxiong Huangkeng town,Guangdong province.In the verification test of the small area,the comparison experiment of the standard nitrogen treatment with the remote sensing model was established,and the results showed that the NDVI of the rice with the standard nitrogen treatment was basically consistent with the nitrogen treatment in the remote sensing model.In the daejeon verification test,the traditional nitrogen application treatment and the remote sensing model nitrogen application treatment were set up to compare the nitrogen treatment the results showed that the application of nitrogen in the remote sensing model was less than that of the traditional nitrogen treatment(the former crop was tobacco or potato),and yield of the remote sensing model nitrogen application treatment was slightly higher than that of the traditional nitrogenous fertilizer treatment.The yield of nitrogen treated by remote sensing model was more balanced.It is indicated that using remote sensing rice variable nitrogen fertilizer decision model to accurately manage rice nitrogen can ensure the yield,reduce nitrogen fertilizer abuse and reduce the input cost of nitrogen fertilizer,and has good application effect.(5)The Decision support system of rice nitrogen nutrition precision management was designed.This system uses the PHP language to programmed,chooses Mapbox as the GIS service engine,the MySQL serves as the database management system,has the planting area information management,the remote sensing image uploads the management,the remote sensing image processing,the fertilization model input and the management,the fertilization prescription graph present and so on function.The operation flow of remote sensing data analysis and decision of variable nitrogen application is standardized,which provides a support platform for rice variable nitrogen fertilizer management,and provides service for the users who need it better. |
PDF Full Text Request