| Farmland nutrient management is the key factor for establishing china food production and food security, because china is known as the agricultural country. Chinese government has given more attention to monitoring and managing the farmland nutrient. But, with the increase of population and development of the economy, it is becoming the more urgent task to enhance the grain yield. Therefore, how to effectively monitor and manage farmland nutrient, to keep or increase grain yield, to reduce farmland pollution, and to raise farm production efficiency is being an very urgent problem. In order to solve the problem that refers to abuse of fertilizer, low nutrient utilization efficiency and excessive soil nutrient accumulation, this article takes the Fu Yang city as an example, explores the space interpolation prediction problem on the farmland nutrients monitoring sample, and studies the choice idea and mean of optimal accessory factor during making farmland nutrient co-kriging interpolation. And then, this article makes comparison between interpolation results accuracy from the optimal accessory factor co-kriging interpolation. Based on the interpolation results, combination with the soil testing formula fertilization scheme and WEBGIS technology, the author designs and develops the Soil Testing Formula Fertilization Information System, which makes the use of monitoring data of soil nutrients more widely and specifically. The main works and results are as following:1. Relevant factor analysis of farmland nutrient spatial variabilityThere is frequently correlation between soil nutrient and other natures. This article studies the relationship between soil nutrient and terrain factors, such as elevation, slope and aspect. The findings indicate that it exists the obviously correlation between soil nutrient and terrain factor. Especially, there are more significant relationship between the soil organic matter and the value of slope’s (aspect) cosine, between rapidly available potassium and the value of aspect’s sine, between total nitrogen and the value of slope’s (aspect). This article makes research on the relationship between soil microelement and soil nutrient. The results suggest that there is more significant correlation between soil nutrient elements and microelement. The correlation between soil nutrient and commutability hydrogen also exists, significant at the0.05level.It also exists relationship among exchangeable aluminum, exchangeable magnesium, exchangeable calcium, effective state manganese, effective state zinc, effective state molybdenum and soil total nitrogen, respectively significant at the0.05level. There was significant correlation between Soil total nitrogen and effective state zinc under the confidence level of0.10. The correlations among commutability hydrogen, effective state of manganese, and effective state of zinc, effective state of molybdenum and soil available phosphorus are significant; the confidence levels are both0.05. There are still correlations between soil available potassium and exchange hydrogen, exchange aluminum, effective state of manganese, effective state of zinc, effective state of molybdenum, significant at0.05levels, respectively. In addition, this article studies the correlations between soil nutrients data. The results showed that there is general significantly relationship among four soil nutrient elements and the confidence level is0.01. The coefficient is0.932between Soil organic matter and total nitrogen content. Moreover, there are high negative correlation between the soil PH value and effective phosphorus, rapidly-available potassium.2. The optimal factor’s cokriging interpolation of farmland nutrients and accuracy testBased on the space autocorrelation theory, combination of accessory factor’s co-integration relationship, cokriging interpolation can be used to predict variable in space view. And the predict accuracy is superior to others. Meanwhile, there are many associated factors with soil nutrients elements, which are as accessory variable for interpolation. Therefore, this article firstly explores the correlation between soil nutrient elements and terrain, soil microelement, and the correlation among soil nutrient elements. According to the strength of the correlation between related factor and different soil nutrient elements, stronger correlation factors are prior to be selected as a cofactor for Cokriging interpolation, and test optimal accessory factor interpolation results by the means of cross test and inspection points. Then, it makes two comparisons. The first accuracy comparison is between the results from soil nutrient’s ordinary kriging interpolation and the results from the suboptimal accessory factor kriging interpolation. The second accuracy comparison is between the results from soil nutrient’s ordinary kriging interpolation and the results from the optimal accessory factor kriging interpolation. The results from cross test and inspection points test illustrate that optimal accessory factor kriging interpolation is superior to suboptimal accessory factor and general kriging. And the predicted accuracy has been improved obviously. Thus optimal accessory factor kriging interpolation provides precision guarantee of application of soil nutrient data.3. The space-time evolution of farmland nutrient analysis and evolution in the soil testing soil fertilizer application. This article exploring the optimal cofactor of kriging interpolation method of soil nutrients, on the basis of applying this method to multiple time points (2004,2006,2008,2010and2011), the difference of soil nutrient value, get the soil nutrient prediction results of multiple time points, and then, based on the study of spatial and temporal variation characteristics of soil nutrients, summing up the space-time evolution of soil, found the study of four soil nutrient elements are faster and faster pace of change in the change characteristics of the four nutrient elements at the same time there exist certain differences in the evolution of:organic matter and total nitrogen of amplitude is relatively small, and the effective phosphorus and available potassium luffing relatively large. In addition it is worth noting that the organic matter, total nitrogen, the biggest change in effective phosphorus and available potassium although the difference is very big, but it changes the most concentrated area exists in low range. This article also from the soil type, land use mode and terrain factors from three aspects, analyzes the spatial and temporal variation characteristics of soil nutrients, soil type and the relationship between spatial and temporal variation of soil nutrients is not obvious; Land utilization way and the relation between spatial and temporal variation of soil nutrients is close, in the area of human activity influence, nutrient variation faster; Terrain factors and nutrients also closely tied to space and time variation, nutrient variation severe areas tend to have plain mountain transition, such as river upstream areas of complex topography. In addition, based on the soil of fuyang soil testing and formula fertilization scheme of soil classification standard according to the different time scales, the research of soil total nitrogen, available phosphorus, available potassium content evolution characteristics, found in the short term (e.g., two years) the absolute change of the soil total nitrogen content reached the proportion of soil classification threshold is low, along with the increase of time scale, the proportion of total nitrogen variation over the threshold will have obvious rise; And the absolute variation of effective phosphorus and available potassium content reached the proportion of soil classification threshold has been high. On this basis, this paper suggested that the daily monitoring of total nitrogen content in soil nutrient monitoring frequency and number of sampling points may be appropriately under monitoring of soil effective phosphorus and available potassium.4. Design and implementation of soil testing and formula information management systemBased on the theory analysis and case study, this article assigns the result of optimal accessory factor kriging interpolation to cultivated land figure spot of Fu Yang present landuse map, combination of geographical data, builds the GIS space database. Further, many kinds of GIS map service and geo-processing service have been built and published by the use of ArcGIS Server. Following, the soil testing and formula information management system can be accessed on the internet. Through Microsoft IIS, this article manages, publishes map service, and provides interface. Eventually, the task of monitoring soil nutrient spot and continuous interpolation within soil nutrient area can be accomplished based on web. The soil testing formula fertilization information management system has been entered into the stage of widely used in Fu Yang agriculture department. |
