土壤质量评价指标体系的构建及评价方法

李鑫,张文菊,邬磊,任意,张骏达,徐明岗

Advance in Indicator Screening and Methodologies of Soil Quality Evaluation

LI Xin,ZHANG WenJu,WU Lei,REN Yi,ZHANG JunDa,XU MingGang

表2 土壤质量评价方法的优缺点

Table 2 Advantages and disadvantages of the soil quality assessment method

评价方法 Assessment method 原理 Principle 优点 Advantage 缺点 Disadvantage 参考文献 Reference 多变量指标克立格法 Multiple-variable indicator kring method 利用多变量指标转换生成土壤质量结合GIS完成评价 Using multiple-variable transformation to integrate soil quality combined with GIS to complete the evaluation 自动评价、动态监测,扩展评价范围 Automatic evaluation, dynamic monitoring, expanding the scope of evaluation 运算过程复杂,模型选择缺乏依据 Complicated calculation process, lack of basis for model selection [8], [34] 土壤质量动力学方法 Dynamics method 利用系统动力学的方法描述土壤质量的变化 Using system dynamic method to describe changes in soil quality 可动态、持续监测土壤质量 Dynamic and continuous monitoring of soil quality 无法进行大空间尺度下土壤质量的比较 Unable to compare soil quality at large spatial scales [5] 土壤质量综合评分法 Integrated scoring method 通过生产量、土壤侵蚀量、地下水、地表水、大气和食物质量评价土壤质量 Evaluating soil quality through production, soil erosion, groundwater, surface water, air and food quality 可全方位、多角度评价土壤质量 Comprehensive and multi-angle evaluation of soil quality 权重影响因素较多,计算较为 困难 Difficult to calculate weight with numberous factors influencing weight [35] 土壤相对质量法 Relative quality method 以假设的理想土壤为标准,评估研究区土壤的相对质量 Evaluating relative quality of the soil in the study area with hypothetical ideal soil as the standard 可定量评估土壤质量,能够进行比较 Quantitative assessment of soil quality, able to compare 需选择理想土壤,若选择不当造成评价偏差 Evaluation bias caused improper selection of ideal soil [8], [36] 土壤质量指数法 Soil quality index method 针对土壤功能选取指标,将指标定量化并综合生成土壤质量指数 Quantifying indicators and integrating a soil quality index selected for soil function 适用于土壤可持续管理,应用范围较广 Suitable for sustainable soil management with a wide range of applications 评价过程中部分受主观影响,易出现偏差 Partly influenced by subjectivity, prone to deviation [25], [37] 灰色关联分析法 Grey relational analysis method 通过关联度和关联序反映各评价对象与理想对象的接近次序,对评价对象进行比较 Evaluating close order and comparing the evaluation objects between objects and ideal object by correlation degree and order 方法简便,评价结果客观 Easy to get an objective evaluation result 若指标值离散,会丢失部分信息 Prone to lose information with discrete values of indicators [38] 物元法 Matter-element method 以物元理论和可拓集合理论为基础,建立多指标性能参数的质量评定模型 Establish a quality evaluation model for multi- index performance parameters based on matter- element and extension set theory 解决单项指标评价的不相容性 Solving the incompatibility of individual index evaluation 计算过程相对复杂 Relatively complicated calculation process [39] 人工神经网络法 Artificial neural network method 通过样本训练使模型进行自学习、自适应,再利用模型对土壤质量进行评价 Training the model for achieving self-learning and adaptive through samples to evaluate soil quality 自动化程度高,避免主观干扰 Avoiding subjective interference with high degree of automation 若学习样本选取不当,易造成评价偏差 Evaluation bias caused improper selection of learning samples [40] TOPSIS法 TOPSIS method 计算样本与最优解和最劣解的相对距离来评价样本的优劣 Calculating the relative distance between samples and the optimal/worst solution to evaluate the quality of samples 排除主客观权重导致的误差 Eliminating errors caused by subjective and objective weights 量化指标较难,灵敏度不高 Difficult to quantify the index, low sensitivity [41]