土壤分类专家系统设计

土壤系统分类以定量化指标划分土壤类型,为土壤自动分类系统的建立提供了可能.为了模仿专家的思维过程实现土壤自动分类,该文着重研究了如何采用面向对象的方法表达专家经验规则、诊断层和诊断特性知识,在此基础上提出了土壤分类决策树的构造方式和推理方法;并从土壤系统分类专家的知识表示和推理机的构建两个方面,提出了土壤分类专家系统 (SCES)的体系结构.GIS作为管理空间信息和提供空间分析方法的平台,也被有效地集成到系统中,以实现土壤自动分类的目的.     

基于GIS的中国土壤分类专家系统设计

土壤系统分类以定量化指标划分土壤类型,为土壤自动分类系统的建立提供了依据;地理信息系统(CIS)的应用扩展了土壤空间信息表达的能力,为实现土壤分类自动制图提供了手段。本文着重模仿专家的思维过程,探讨如何实现土壤自动分类。(1)采用面向对象的方法表达专家经验规则、诊断层和诊断特性知识,并在此基础上提出了决策树的构造方式和推理方法;(2)从土壤系统分类专家的知识表示和推理机的构建两个方面,提出了土壤分类专家系统(SCES)的体系结构、土壤自动分类与地理信息融合的方法和基于GIS和专家系统(ES)的土壤自动分类检索系统的框架结构。     

以中国土壤系统分类为基础的土壤参比

为了土壤信息交流、知识共享 ,作者首先阐述了与世界土壤分类发展相伴随的中国土壤系统分类的发展 ,进而对中国土壤系统分类与美国土壤分类和WRB方案进行了参比 ,然后着重讨论了它与土壤发生分类之间的参比。今后拟建立土壤参比信息系统以实现计算机自动参比     

中国土壤系统分类在浙江省平原旱地土壤分类中的应用

本文应用《中国土壤系统分类（修订方案）》的诊断层和诊断特性，探讨了浙江省平原旱地土壤在中国土壤系统分类中的归属，并比较了土壤系统分类和发生了分类和发生分类对本区土壤划分的异同，最后提出了一些建议。     

梅花山土壤诊断特性及系统分类研究

鉴定了梅花山自然保护区土壤的诊断诊断层和诊断特性，分析对比了本区土壤在地理发生分类制和中国土壤系统分类制中的类型归属，初步探讨了中亚热带湿润地区山地土壤系统分类类别的垂直分布规律，并就本研究供试土壤涉及到的问题对中国土壤系统分类是提出修订建议。     

对《中国土壤系统分类(首次方案)》的几点建议

《中国土壤系统分类(首次方案)》,将我国原来以地理发生观点制定的土壤分类系统,初步改为以"诊断层"、"诊断特性"为依据的新分类系统,从而与国际土壤分类学界有了共同语言。为使系统分类更加完善,作者提出了5个方面的建议。     

武夷山土壤分类参比

按照《中国土壤系统分类（修订方案）》鉴定了武夷山垂直带中６个代表性剖面的诊断层,诊断特性和系统分类名称,并与发生学分类。美国土壤系统分类以及世界上土壤资源参比基础等不同分类系统中的土壤类别归属作了对比研究,为中国土壤系统分类在我国东南部山地丘陵区的应用提供示范。     

云南省元谋干热河谷的土壤系统分类

应用土壤系统分类的原理和方法，研究了元谋干热河谷主要土壤的土壤铁铝特性、弱腐殖质化（层）特性、变性特征、岩性特征等诊断特性。据此提出了当地高级分类单元的土壤系统分类（草案）。分析了主要土壤的剖面特征和理化性状，并对亚类以下分类问题提出了看法。     

中国南方富铁铝化土壤土系建立进展

土壤基层分类单元划分，特别是土系建设是土壤系统分类的基础，也能直接为农业生产提供更详细的科学指导。富铁铝化土壤（砖红壤、赤红壤和红壤）在我国南方广泛分布，为了解其在中国土壤系统分类中归属以及土系建设情况，本文综述了自2008年以来土系研究的相关成果，系统分析了富铁铝化土壤诊断层和诊断特性检索、高级分类单元归属以及基层分类单元建设现状，并展望了土系建设的发展方向：进一步细化土系划分和描述标准，加强新土系建设，加强新技术和新方法在土系建设中的应用，加大和扩展土系建设成果的运用，旨在为我国南方富铁铝化土壤系统分类研究提供参考。     

海南岛土壤发生分类类型在系统分类中的归属

本文对海南岛 2 3 7个土壤剖面 ,按中国土壤系统分类体系 ,逐一进行鉴别、检索、分类定名 ,阐述按土壤发生分类体系的 2 8个亚类级土壤类型在中国土壤系统分类中的归属 ,提出按中国土壤系统分类的该岛土壤分类系统建议 ;并讨论了该岛土壤发生分类与系统分类之间具体土壤类型的对应关系。     

The problem of soil interpretation according to the WRB 2014 classification system in the context of anthropogenic transformations

Evaluation of soil transformations and degradation is important not only for agriculture, but also for fundamental research on identification of new features for classification. In 1999, new Lithuanian soil classification (LTDK-99) was developed by adopting the diagnostic principles of WRB 1998. Together with new knowledge, it highlighted some relevant soil research and classification problems. Not only new terms but also diagnostic features of several newly distinguished first-level soils groups are being discussed, especially for the identification of cambisols, planosols, arenosols, and regosols. Additionally, a group of stagnosols is suggested for inclusion into this classification. The presence of glaciogenic secondary clay minerals in Lithuanian cambisols does not imply unambiguous brownification involvement, neither any study shows cambisols having specific to brownification iron minerals. The intensively occurring lesive in these soils questions correctness of identification of cambisols additionally aggravated by strong and deep anthropogenization of the soil profiles within the territorial range. Complex glaciogenesis makes it difficult to identify Lithuanian planosols. WRB 2014 (update 2015) identifies planosols according to the pedogenesis in old, strongly weathered binomial deposits using examples mostly from South American, South and East African plateaus of the subtropical zone. The same classification assigns planosols identified in Lithuania to stagnosols, that is, to the soils characterized by binomial genesis and stagnification. The diagnostic features of arenosols and regosols are similar in both groups and are related to primitive pedogenesis. Difficulties in distinguishing between the two groups in Lithuania are strengthened by the intensive cultivation of the ecosystems in their territorial range. The existing problems in Lithuanian soil diagnostics underline the questions of soil classification in general, especially regarding the evolution tendencies in soil classification. The article draws attention to the soil diagnostic aspects like distribution and genesis of clay particles, genesis of territory, and anthropogenic transformation for the improvement of soil identification.     

天山中部北麓丘陵地带土壤发生特性与系统分类

通过样地调查与土壤理化性质分析相结合对天山中部北麓丘陵地带的土壤特性进行研究,确定各剖面的诊断层和诊断特性及其在系统分类中的位置。结果表明天山中部北麓丘陵地带在系统分类中可划归为3个土纲4个土类;系统分类中土壤的分类位置与发生分类位置并不具有简单的一一对应的关系。     

红壤区土壤资源利用与管理信息系统的研制

研究建立土壤资源信息系统,实现土壤资源定位管理,是合理开发我国南方大面积红壤资源的前提。以浙江省红壤区为例,设计的省、县、农场三级管理层次的土壤资源利用与管理信息系统（ＵＭＳＩＳ）包括基础数据库和评价、规划、咨询三个主模型库,具有数字化输入、图象处理、数据库管理、综合制图、空间相关分析等功能。目前已建立的浙江省、衢州市和龙游县三个不同比例尺实验性系统,已初步完成了系统调试和运行,并应用于区域红壤资源的资源分类、适宜性评价、水土保持、农业开发分区等工作。另外,研究也表明商用ＧＩＳ软件和应用模型的有机链接仍然是当前ＧＩＳ二次开发较为现实的途径。     

河南省典型土系的特定土层特征与分类研究

根据土壤发生学理论,在土壤景观相似的土壤中(如相同母质、相似气候、相同地形以及相似的植物类型和动物活动),土壤的属性具有相似性;然而由于地理环境中复杂的成土因素的影响,土壤绝无完全相同的土壤个体。在划分土壤地理发生分类基层单元土种和系统分类基层单元土系时,特定土层被用于反映土体构型差异。但特定土层类型多样,目前尚处于模糊状态,随着土系划分研究的深入,有必要对其进行分门别类与归纳整理。本文以河南省典型土系为例,梳理和总结了土系划分时涉及的主要特定土层,从特定土层形成根源上将其分为现代成土环境下土壤发生过程形成的、地质成因或地质时期土壤环境下发育形成的和人类活动产生的三个类型,并提出了基于土系概念的特定土层分类与检索体系,旨在使特定土层的概念更为系统、全面、开放,以便于使用。     

土壤侵蚀对坡耕地耕层质量退化作用及其评价趋势展望

土壤侵蚀是导致坡耕地耕层质量退化和土壤生产力不稳定的关键驱动因素。该文从水蚀区坡耕地侵蚀控制和生产功能角度,在解析地块尺度土壤侵蚀、水土保持、农业活动对坡耕地耕层生态过程作用特征的基础上,系统分析了土壤侵蚀对坡耕地耕层质量退化作用、影响效应及作用途径。认为:1)坡耕地耕层质量变化由降雨侵蚀、耕作活动交互作用的生态过程决定,2种作用的时间、空间尺度不同;耕层土壤参数在坡耕地农业生产中作用分为保水、保土、保肥和增产潜力,由地块尺度农作物-耕层耦合效应决定土壤生产能力、坡耕地水土流失特征及耕层侵蚀性退化方向及程度。2)土壤侵蚀对坡耕地耕层质量退化作用表现为土壤性质恶化、土壤质量劣化、土地生产力衰退3个方面,耕层土壤物理性质变异程度大于化学性质变异,径流作用导致的土地生产力衰退大于土壤流失作用。3)坡耕地耕层质量评价指标体系应兼顾侵蚀下降、产量提升2个目标,地块尺度诊断指标有效土层厚度、耕层厚度、土壤容重、土壤抗剪强度、土壤有机质、土壤渗透性可作为合理耕层评价最小数据集;坡耕地合理耕层适宜性分为5级,其诊断指标分级标准宜与土壤侵蚀分级和耕地地力分级衔接。4)坡耕地合理耕层评价未来应密切关注耕层质量诊断指标最小数据集、坡耕地合理耕层阈值/适宜值分级标准、坡耕地水土流失阻控标准拟定3个主要方向。研究可为深入认识坡耕地侵蚀性退化机制,辨识坡耕地合理耕层调控途径以及坡耕地合理耕层构建技术参数提供依据。     

Comparison of the results of soil profiles’ diagnostics performed in three classification systems

Three soil classification systems—the World Reference Base for Soil Resources (WRB), Soil Taxonomy, and the recent Russian system—were used for the identification of 17 soil profiles in southwestern Poland; all the systems put emphasis on the soil properties as diagnostic criteria. Different soils developed on glaciofluvial plains, loessic uplands, and in the Sudetes Mountains were classified. The best correlation between the classification decisions in the different systems was obtained for the most widespread soils owing to the similarity of the diagnostic criteria, which were essentially close although not coinciding. The most prominent divergence between the systems in both the names and the taxonomic categories of the soils was found for the polygenetic soils and for the soils developing from the lithologically discontinuous parent materials. It was also found that the diagnostic elements differ in terms of their taxonomic importance among the classification systems.     

A BASIS FOR THE CLASSIFICATION OF SOIL

Considering the nature of soil and the ends which a classification must serve, principles are stated whereby a soil classification may be devised for application over extensive areas of varied soil composition. Naturally occurring bodies of soil, each with a high degree of homogeneity, are apparent rather than real individuals as their properties overlap to form a continuum. This continuum is multi-dimensional because soil is characterized by numerous properties. The procedure of devising a classification is one of subdividing the continuum such that class boundaries accommodate, as far as possible, apparent individuals rather than of grouping like apparent individuals together. A classification may be used to locate the position of a profile in the continuum and so define its relationship with other profiles. It may also be used to indicate the soil composition of land by using soil classes on a map to show differences in the soil mantle. The latter procedure may best be regarded as land classification or soil mapping rather than soil classification; a class of land or mapped area seldom contains profiles belonging exclusively to a single class, whereas a soil class never contains profiles of another class. The soil form, a specific arrangement of diagnostic horizons, is introduced as a category above the series to facilitate the identification of soil profiles. Member series of a form are defined according to property variations within the diagnostic horizons of the form. It is suggested that a binomial system of profile nomenclature, using the form and the series, would have much to commend it.