| This study aimed at Xinjiang production and construction corps cotton production has characteristics of vast farmland area, high mechanization degree, organization and management is relatively uniform and soil nutrient difference of continuous farmland is strong, integrated with GIS, partition management and nutrient diagnosis technology, combined with geo-statistics and multivariate statistical analysis, research the effect of different soil sampling, sampling interval and spatial interpolation method on drip irrigation cotton. Based on different conditions of the content of soil nitrogen nutrient status, distribution under different nitrogen nutrient partitions, this paper researched soil nitrogen partitioning precision influence to determine the actual distribution of the high precision fitting cotton nitrogen optimal sampling, sampling interval and spatial interpolation method; We adopted high precision soil sample collect method and spatial interpolation technique, divided study area into high, medium and low three regions by nitrogen status, through conducting different nitrogen levels fertilization experiments, finally established base fertilizer and topdressing recommended fertilization models according to soil alkali-hydrolyzable nitrogen and NDVI for diagnostic indicators. Through the research of variable rate fertilization application based on basal recommended model, compared the change of soil nitrogen, cotton growth, yield and other indicators between variable fertilization and traditional fertilization, verified the effect of variable rate fertilization; By monitoring the change of cotton leaf NDVI value in different partitions, carried out on NDVI index drip irrigation cotton fertilizer recommendations, and got the effect of model verification by field application. In short, this paper carried out high precision n partition management and fertilization model, validation and application systematically, provided basic gist for fertilization by soil, soil quality improvement and promotion and cultivated land ecological optimization, also offered technological support for further explore suitable soil nitrogen nutrient division management model for production and construction corps agricultural production and operation. The main research conclusions are as follows:1. Study of the soil nitrogen nutrient management partition method optimizationVariation values of soil organic matter, total nitrogen, alkali solution nitrogen, available phosphorus and available potassium nutrient in the study area are between 35.27% and 56.63%, which conform to the basic conditions of the precision agriculture management zoning and variable rate fertilization implement. Through correlation analysis between soil attribute data and soil total nitrogen, we found that organic matter and alkali solution nitrogen were positive correlation with total nitrogen. This study selected total nitrogen, organic matter and alkali solution nitrogen three factors which had high correlation with soil nitrogen as evaluation index of nitrogen nutrient division, and divided the management partition; Through the semivariance function analysis of soil nitrogen we found that the best fitting model of total nitrogen is spherical model, while organic matter and alkali solution nitrogen are index model; nugget effect of total nitrogen, organic matter and alkali solution nitrogen is 25%<(C0/C0+C)<75%; Through the contrast between the different prediction methods we found that the best index optimum space forecast method Kriging interpolation method interpolation precision is the highest on soil total nitrogen and organic matter, alkali-hydro nitrogen nutrient, then the LPI. So the best interpolation method in the study was Kriging interpolation; Through analysis of soil total nitrogen, organic matter and alkali solution nitrogen under different sampling methods it showed that the reaction of distribution patterns of nutrients sampling method OK in the interpolation is tiny, from the independent test results we found that the nutrient ME absolute value was more close to 0 under grid sampling method than that under random sampling method, and the RMSE of three kinds nutrient under random sampling method is relatively bigger than that under grid sampling method. Therefore, we concluded that interpolation results of soil total nitrogen, organic matter and alkali solution nitrogen nutrient under grid sampling method is more accurate. And through comparing the test values and the measured values of study area we saw that the grid sampling method is the best sampling method; Comprehensive analysis the interpolation error and independent inspection and nutrient map of soil total nitrogen, organic matter and alkali solution nitrogen under different sampling interval we found: in the study area, sampling interpolation precision of total nitrogen, organic matter and alkali solution nitrogen with 20 m sampling space is the highest, though the precision with 30 m sampling spacing is worse than that with 20 m sampling spacing, but consider of actual work such as sampling, test, human and material investment, this paper adopted 30 m sampling spacing to conduct nitrogen nutrient division management in the variable rate fertilization application.2. Research of precision nitrogen partition and fertilization model of drip irrigation cottonUsing the optimal sampling, the sampling spacing and spatial interpolation method, the study area is divided into three precise nutrient division(low, medium, high); Three kinds of soil nutrients of three partitions has spatial autocorrelation, membership’s best half variance fitting model of three partitions are globular respectively(Spherical), Gaussian(Gaussian) and index(Exponential) model, and space variation memberships degree of partitions are more than 20 m, once again proved that 20 m spacing for the best sampling spacing and the partition number is reasonable; Nitrogen content of partition 3 is the highest, then partition, partition 1 is the lowest; We build up regression model between nitrogen fertilization and cotton yield, we concluded the highest seed cotton yield 6305 kg respectively, hm-2, 6762 kg, hm-2 and 6794 kg, hm-2, corresponding to the nitrogen application rate of 391 kg, hm-2, 305 kg, hm-2and 280 kg, hm-2; We use the nitrogen content when gain then maximum yield as the best nitrogen, and with soil alkaline hydrolysis nitrogen of 0-20 cm, 20-40 cm, 40-60 cm, 60-80 cm of low nitrogen, the nitrogen and high nitrogen area as diagnostic indicators to establish basal fertilization system of soil alkali-hydrolysable nitrogen in 0 ~ 20 cm, finally ensure the optimum nitrogen of different partitions; Cotton canopy NDVI value increased gradually with the increase of nitrogen and pretend linear trend, and the determination coefficient were high. By analyzing the cotton NDVI of different growth period and cotton yield, established the cotton yield estimation model of each key growth period, and determine the critical value of NDVI based on LNA and NDVI regression equations. Integrated cotton yield estimation model, LNA and critical diagnostic value of the relationship between NDVI and established drip irrigation cotton partition topdressing model based on NDVI as nitrogen nutrition diagnosis index.3. Research of basal fertilizer variable rate fertilization technology based on high precision nitrogen partitioningWe established the high-precision basal prescription figure of variable rate fertilization, and applied it in the equipment of variable rate fertilization, the variable rate fertilization error is between 0.14% ~ 6.65%, the control of variable rate fertilization under variable machinery operation in the process of the actual shoes and recommended fertilization rate between the control accuracy is above 93%; The average recommended basal fertilizer is 46.86 kg/hm2, through the variable machinery, actual fertilizer rate is 48.49 kg/hm2, compared with the traditional fertilizing 75 kg/hm2,it reduced nitrogen input about 35.35% per hectare; Cotton dry matter accumulation and nitrogen uptake in variable rate fertilization area are significantly or extremely significantly increased than that in traditional fertilizing area, suitable amount of nitrogen fertilizer can significantly increased cotton dry matter accumulation of cotton leaves and nitrogen uptake to some extent, while increase the cotton boll number per plant and single boll weight and finally increase cotton yield.4. Nitrogen partition topdressing model validation of partition and application of drip irrigation cottonThrough application and effect verification of the partition topdressing model according to the model based on nitrogen nutrition diagnosis we found that the standard error of partition recommended fertilization cotton NDVI decreases gradually, and as cotton growth, three partitions tend to accordance gradually; Compared with traditional fertilization, we found that the effect of topdressing recommendation on NDVI in the early growth period of cotton is little, from boll-forming stage the effect become bigger; The standard errors of LAI under partition recommended fertilization in the cotton key growth period were less than the normal processing of corresponding, as cotton growth, standard error of LAI in recommended fertilization area present decreases trend, compared with traditional partition fertilizer we found that effect of topdressing recommendation on LAI was bigger than that of conventional fertilization in low nitrogen nutrient area; Contrast fertilizing amount and yield of each partition recommended fertilization with conventional fertilizer production we found that the amount of the fertilization was low nitrogen area > conventional fertilization area > medial nitrogen area > high nitrogen area; fertilizer quantity of low nitrogen area was higher than that of conventional fertilizer quantity is 15.4 kg·hm-2, fertilizer quantity in high nitrogen area and medial nitrogen area lower than that of conventional fertilizer quantity were 34.1 kg, kg·hm-2 and 53.1 kg·hm-2.When fertilizer application are same, the basic fertility of soil determine the cotton yield, Nitrogen use efficiency of low nitrogen area in the conventional fertilization treatment is the lowest, nitrogen use efficiency of medial nitrogen area is the highest, while the rule of under the nitrogen use efficiency in the recommended fertilizer treatment regularity is opposite; nitrogen use efficiency in the fertilizer recommendation treatment were greater than normal processing, and the with the nitrogen use efficiency in high nitrogen area was the highest. |