粮食产量空间化中4种误差修正方法的对比与分析

发布时间：2018-04-08 07:09

本文选题：统计数据　切入点：空间化　出处：《中国农业资源与区划》2017年08期

【摘要】：[目的]空间化作为一种常用的地学数据处理方法,必然会存在一定的误差,而对空间化结果进行误差修正,可以降低空间化误差。[方法]文章以2005年粮食产量空间化为例,以各地市不同农田类型(水田、水浇地、旱地)面积数据为自变量,以各地市粮食产量统计数据为因变量,进行多元线性回归分析建模。在具体建模时,令常数项为0,将全国分为7个区,各区分别建立回归方程。然后分别利用4种误差修正方法对空间化初步结果进行修正。选取4种误差评价因子,对修正前后的空间化结果的精度进行对比和分析。[结果](1)均值法不能被用于修正空间化初步结果;(2)比例系数法、权重系数法Ⅰ(不同农田类型同一权重)和权重系数法Ⅱ(不同农田类型不同权重)3种方法都可以被用于修正空间化初步结果;(3)利用权重系数法Ⅰ修正后的空间化结果的精度最高,比例系数法次之,权重系数法Ⅱ最差。[结论]误差修正方法对提高空间化精度具有重要影响。该研究虽以粮食产量空间化为例,但所得结论同样适用于其他社会经济统计数据的空间化研究,对以后统计型数据空间化研究具有一定的参考价值和指导作用。
[Abstract]:[Objective] spatialization as a common method of processing data, there must be some errors, but the results of spatial error correction, this method can reduce the error in space.] 2005 grain output space as an example, in different types of farmland around the city (paddy field, irrigated land, dry land area) the data as independent variables and dependent variables in food production around the city for statistical data, multivariate linear regression analysis model. In the model, so that the constant is 0, the country is divided into 7 zones, regression equations were established respectively by 4 districts. Then the error correction method for spatial correction. The preliminary results of selection 4 kinds of error evaluation factors were compared and analyzed. Results the results of space before and after the correction accuracy of the] (1) the average method cannot be used for modification of the space of the preliminary results; (2) proportional coefficient method, weight coefficient of different ( Farmland type the same weight and weight coefficient method (II) for different soil types with different weight) 3 kinds of methods can be used for modification of the space of the preliminary results; (3) using the method of weighting coefficients of the space I modified the highest accuracy, proportional coefficient method of weight coefficient method of the worst Conclusion]. The error correction method has an important impact on improving the spatial accuracy. While the study on spatial grain production as an example, the research space but the conclusion is also applicable to other social and economic statistical data, the spatial data after statistical research has certain reference value and guiding role.

【作者单位】：防灾科技学院;河南大学环境与规划学院;华东师范大学地理信息科学教育部重点实验室;云南大学资源环境与地理科学学院;
【基金】：中国科学院战略性先导科技专项“应对气候变化的碳收支认证及相关问题”(XDA05050000)
【分类号】：F224;F326.11

【相似文献】