【作者】 王宁宁；

【导师】 和钟铧； 李晓燕；

【作者基本信息】 吉林大学 ， 土地资源管理， 2009， 硕士

【摘要】 21世纪,保障粮食安全是我国农业现代化的首要任务,而实现粮食安全的关键和前提是摸清我国耕地资源的数据和质量。土壤养分是耕地地力的重要标志,研究土壤养分的空间分布是调整管理措施和各种物质投入量、获得最大效益的基础。本文的研究目的是对九台市土壤养分的空间分布进行预测,以便于有效地指导农业生产,为未来推广测土配方施肥,实现精准农业提供一个基本保障;为实现农业可持续发展提供重要的理论和实践意义。在研究中,选择了以往常用的普通克吕格法和反距离权重法以及近年来比较流行的回归克吕格法三种方法。通过研究发现,回归克吕格法虽然近年来比较流行,但由于九台市土壤养分的影响因素过于复杂,并不适合九台市土壤养分的空间分布预测研究。普通克吕格法和反距离权重法对于土壤养分分布趋势的反映上基本是一致的,反距离权重插值的精度要高于普通克吕格插值的精度,说明九台市土壤养分的空间分布预测研究更适合采用反距离权重方法。最终的研究结果表明,有机质和速效钾的分布规律比较相似,呈现出由西南方向向东北方向逐渐减少的趋势。但是有机质的较高值主要出现在作物为玉米的黑土上,而速效钾的高值主要出现在作物为水稻的草甸土上。速效磷的分布是由南向北逐渐递减的,高值主要出现在台地地区的草甸土上,耕地类型为旱田,作物类型主要为玉米。全氮在整个研究区的分布普遍较低,最高值零落的出现在中南部,西南部等附近台地地区的黑土和草甸土上,主要粮食作物有玉米、蔬菜和水稻。更多还原

【Abstract】 In recent years, the spatial variability of soil nutrients have been widespread concern and attention with the protection of food security and the rapid development of precision agriculture. According to the research status at home and abroad, the paper chose the Ordinary Kriging, Regression Kriging and Inverse Distance Weighted Interpolation methods to analysis soil nutrients spatial variability characteristics and distribution pattern in Jiutai County. Studies had shown that because of soil nutrient factors that affected differences in the larger, different nutrients were required in different directions. Study results had a great significance to reveal the spatial variation of soil nutrients in the law, as well as the precise and accurate adjustment of soil management measures to optimize the use of nutrients to maximize resources, to obtain the maximum yield and maximum economic efficiency and to reduce the negative impact on the environment, toprotect the land and other agricultural natural resources. The main conclusions were as follows:The variation coefficient of Four types of soil nutrients (Total N, Organic matter, Available P and Available K) was between 10-100%, belonging to the middle variation. These changes were as a result of structural factors and random factors. The normal distribution of data found that the four original data did not comply with normal distribution. The only Total N was not in line with the normal distribution after log transfermation, So it can not use Ordinary Kriging.To choose some representative environmental factors (terrain factor and NDVI) analysised their correlation with soil nutrients in Jiutai City by ArcGIS, four types of soil nutrients at the level of 0.01 showed negative correlation; Organic matter and Available K at the level of 0.01 showed negative correlation with slope and terrain relief; Organic matter and Available P at the level of 0.01 showed negative correlation with surface roughness; four types of soil nutrients related to vegetation (NDVI) were positive correlated. Although the soil nutrients and environmental factors demonstrated a certain degree of relevance, by fitting multiple linear regression equation found that the agreed coefficient was too low that it lost the significance. Regression Kriging method is more popular now as a prediction of soil nutrients, but the factors which impact it were too complex. However it was not fit to use Regression Kriging fou soil nutrients in Jiutai City.The ratio of nugget value and base value about Organic matter and Available K was about 20%, showing a strong spatial correlation,that noted structural factors (soil parent material, topography, climate and other natural factors) played a major role to the spatial variability of this two types of soil nutrients . The ratio of nugget value and base value about Available P wsa about 70%, showing a medium spatial correlation, that noted random factors (such as cropping systems, fertilization, cultivation system) played a major role to the spatial variability of available P.The prediction map was basically consistent of Ordinary Kriging and Inverse Distance Weighted Interpolation methods to soil nutrients-Organic matter, Available P and Available K. The distribution of Organic matter and Available K were basically the same, showing a direction from southwest to northeast gradually decreasing trend. However, the higher data of Organic matter mainly was in the black soil of corn crops, but the higher data of Available K occurred mainly in the rice crop for the meadow territories. Jiutai City, northeast of the high-lying, southwest low-lying, topographic factors on the distribution of organic matter and potassium played an important role in the control. The distribution of Available P was gradually descending from south to north, and high-value mainly occurred in the tableland areas meadow soil above of which crops mainly maize and arable land types is glebe. Inverse Distance Weighted Interpolation diagram showed that Total N in the distribution of the entire study area was generally low and the maximum value emerged in the lower of part of the region and the southwest and other areas near the mesa and meadow black earth soil above which the main food crops of corn, vegetables and rice grew. This may be the common results of intrinsic factors such as the topography and external factors such as man-made systems, fertilization in Jiutai, however human economic activity made the distribution of nutrients tend homogenization.To comparison Kriging interpolation and Inverse Distance Weighted Interpolation found that Inverse Distance Weighted Interpolation was more suitable to forecast the spatial distribution pattern of soil nutrients in Jiutai City, and its accuracy was higher than Kriging interpolation, especially Available P. This might be related to the environment of Jiutai City itself. So it was more suitable for using Inverse Distance Weighted Interpolation to forecast spatial variability of soil nutrients in Jiutai City.更多还原