基于GIS农田土地质量评价与立地分析——以京郊房山区良乡为例

The Quality Evaluation and Site Assessment of Agricultural Land Based on GIS:A Case Study in Liangxiang Town, Fangshan District, Beijing

为揭示农田利用可行性 ,提出农田立地概念 ,建立农田评价体系 ,以京郊房山区良乡为例 ,基于GIS技术评价农田质量 ,分析农田立地 ,综合评价农田 ,为划定基本农田提供客观依据 ;根据农田评价评分数值分布 ,通过模糊稀疏度算法模型 ,实现农田自动分等定级 ,减少了界定农田等级指标的主观性。

Agricultural land evaluation is defined as evaluating land resources suitability for crop growth and analyzing the feasibility of land use for agriculture. Agricultural land evaluation system is perfected based on quality and site of farmland in order to be accustomed to the farmland protection. This system is designed as a gradation framework which includes ⑴the suitable types of farmland, ⑵the suitable orders of farmland,⑶the value classes of farmland use, and ⑷land resources units. In this paper, the evaluating unit is farmland plot which is used in town area and shown on land use map with one to ten thousands scale. Corresponding to the different gradation of agricultural land evaluation system, the different factors evaluating farmland are chosen. In the light of the above mentioned agricultural land evaluation system, with Liangxiang, a town of Fangshan District, Beijing, as an example, the operating pattern evaluating the suitable types and orders of farmland is put forward based on GIS technique. Firstly, according to the national criteria of farmland fertility drafted by China Ministry of Agriculture and the land evaluation framework published by FAO, the agricultural land quality reference frame is established, which is composed of four kinds of factors including soil fertility, environment condition, engineering and management of farmland. Then farmland units which is fit to cultivation are separated by “Reselect' command of PC ARC/info based on the land quality reference frame. Secondly, the selected farmland units are evaluated further. The farmland quality is evaluated by employing the fuzzy closeness algorithmic technique, and the farmland site is assessed adopting the model of factor scoring sum, then the possibility and feasibility of agricultural land use are revealed, and the prime farmland is confirmed. The data of farmland quality factors come respectively from the datum of Fangshan soil general survey, Liangxiang topographic map and land use map. The power value of selected factor is determined by using the Analytic Hierarchy Process (AHP). Finally, according to the value distribution of farmland total score, the farmland grades are classified automatically by comparing the values between the fuzzy sparseness and closeness, which is realized by Visual Basic Programs. Based on the datum distribution patterns in Liangxiang farmland evaluation maps, some conclusions are shown as the follows: ⑴The values distribution patterns of farmland quality are decided by the geographic distribution of soil types. The evaluated values of farmland quality are high, which is from 0.81 to 0.97(Fig.1). It is fully proved that the farmland is excellent in Liangxiang. ⑵The values distribution patterns of farmland site show that the site high value is scattered in the northwestern area of Jing-Shi highway and the site low value is distributed in the developing area of Liangxiang town planning (1996-2010). The variation range of farmland site values is large, which is from 0.04 to 0.87(Fig.2). And the distribution patterns of Liangxiang farmland summing up values are similar to that of farmland site values. The variation range of farmland summing up values is from 0.21 to 0.90(Fig.3). ⑶Compared with the distribution of Liangxiang prime farmland (1996-2010)(Fig.4) which was drawn by Fangshan House–Land Administrative Bureau, the datum distribution patterns of the farmland evaluation summing up values in Liangxiang is acceptable. The conclusions of Liangxiang farmland evaluation are practicable.