The Ecological Metadata Language (EML) is a metadata standard developed by the ecology discipline and for the ecology discipline. It is based on prior work done by the Ecological Society of America and associated efforts (Michener et al., 1997, Ecological Applications). EML is implemented as a series of XML document types that can by used in a modular and extensible manner to document ecological data. Each EML module is designed to describe one logical part of the total metadata that should be included with any ecological dataset.

To provide the ecological community with an extensible, flexible, metadata standard for use in data analysis and archiving that will allow automated machine processing, searching and retrieval.

The architecture of EML was designed to serve the needs of the ecological community, and has benefitted from previous work in other related metadata languages. EML has adopted the strengths of many of these languages, but also addresses a number of short-comings that have proved to inhibit the automated processing and integration of dataset resources via their metadata. The following list represents some of the features of EML: Modularity: EML was designed as a collection of modules rather than one large standard to facilitate future growth of the language in both breadth and depth. By implementing EML with an extensible architecture, groups may choose which of the core modules are pertinent to describing their data, literature, and software resources. Also, if EML falls short in a particular area, it may be extended by creating a new module that describes the resource (e.g. a detailed soils metadata profile that extends eml-dataset). The intent is to provide a common set of core modules for information exchange, but to allow for future customizations of the language without the need of going through a lengthy 'approval' process. Detailed Structure: EML strives to balance the tradeoff of too much detail with enough detail to enable advanced services in terms of processing data through the parsing of accompanied metadata. Therefore, a driving question throughout the design was: 'Will this particular piece of information be machine-processed, just human readable, or both?' Information was then broken down into more highly structured elements when the answer involved machine processing. Compatibility: EML adopts much of it's syntax from the other metadata standards that have evolved from the expertise of groups in other disciplines. Whenever possible, EML adopted entire trees of information in order to facilitate conversion of EML documents into other metadata languages. EML was designed with the following standards in mind: Dublin Core Metadata Initiative, the Content Standard for Digital Geospatial Metadata (CSDGM from the US geological Survey's Federal Geographic Data Committee (FGDC)), the Biological Profile of the CSDGM (from the National Biological Information Infrastructure), the International Standards Organization's Geographic Information Standard (ISO 19115), the ISO 8601 Date and Time Standard, the OpenGIS Consortiums's Geography Markup Language (GML), the Scientific, Technical, and Medical Markup Language (STMML), and the Extensible Scientific Interchange Language (XSIL). Strong Typing: EML is implemented in an Extensible Markup Language (XML) known as XML Schema, which is a language that defines the rules that govern the EML syntax. XML Schema is an internet recommendation from the World Wide Web Consortium, and so a metadata document that is said to comply with the syntax of EML will structurally meet the criteria defined in the XML Schema documents for EML. Over and above the structure (what elements can be nested within others, cardinality, etc.), XML Schema provides the ability to use strong data typing within elements. This allows for finer validation of the contents of the element, not just it's structure. For instance, an element may be of type 'date', and so the value that is inserted in the field will be checked against XML Schema's definition of a date. Traditionally, XML documents (including previous versions of EML) have been validated against Document Type Definitions (DTDs), which do not provide a means to employ strong validation on field values through typing. There is a distinction between the content model (i.e. the concepts behind the structure of a document - which fields go where, cardinality, etc.) and the syntactic implementation of that model (the technology used to express the concepts defined in the content model). The normative sections below define the content model and the XML Schema documents distributed with EML define the syntactic implementation. For the foreseeable future, XML Schema will be the syntactic specification, although it may change later.

The following section briefly describes each EML module and how they are logically designed in order to document ecological resources. Some of the modules are dependent on others, while others may be used as stand-alone descriptions. This section describes the modules using a "top down" approach, starting from the top-level eml wrapper module, followed by modules of increasing detail. However, there are modules that may be used at many levels, such as eml-access. These modules are described when it is appropriate.

The eml module is a wrapper container that allows the inclusion of any metadata content in a single EML document. The eml module is used as a container to hold structured descriptions of ecological resources. In EML, the definition of a resource comes from the The Dublin Core Metadata Initiative , which describes a general element set used to describe "networked digital resources". The top-level structure of EML has been designed to be compatible with the Dublin Core syntax. In general, dataset resources, literature resources, software resources, and protocol resources comprise the list of information that may be described in EML. EML is largely designed to describe digital resources, however, it may also be used to describe non-digital resources such as paper maps and other non-digital media. In EML, the definition of a "Data Package" is the combination of both the data and metadata for a resource. So, data packages are built by using the <eml> wrapper, which will include all of the metadata, and optionally the data (or references to them). All EML packages must begin with the <eml> tag and end with the </eml> tag. The eml module may be extended to describe other resources by means of its optional sub-field, <additionalMetadata>. This field is largely reserved for the inclusion of metadata that may be highly discipline specific and not covered in this version of EML, or it may be used to internally extend fields within the EML standard.

The eml-resource module contains general information that describes dataset resources, literature resources, protocol resources, and software resources. Each of the above four types of resources share a common set of information, but also have information that is unique to that particular resource type. Each resource type uses the eml-resource module to document the information common to all resources, but then extend eml-resource with modules that are specific to that particular resource type. For instance, all resources have creators, titles, and perhaps keywords, but only the dataset resource would have a "data table" within it. Likewise, a literature resource may have an "ISBN" number associated with it, whereas the other resource types would not. The eml-resource module is exclusively used by other modules, and is therefore not a stand-alone module.

The following four modules are used to describe separate resources: datasets, literature, software, and protocols. However, note that the dataset module makes use of the other top-level modules by importing them at different levels. For instance, a dataset may have been produced using a particular protocol, and that protocol may come from a protocol document in a library of protocols. Likewise, citations are used throughout the top-level resource modules by importing the literature module.

The eml-dataset module contains general information that describes dataset resources. It is intended to provide overview information about the dataset: broad information such as the title, abstract, keywords, contacts, maintenance history, purpose, and distribution of the data themselves. The eml-dataset module also imports many other modules that are used to describe the dataset in fine detail. Specifically, it uses the eml-methods module to describe methodology used in collecting or processing the dataset, the eml-project module to describe the overarching research context and experimental design, the eml-access module to define access control rules for the data and metadata, and the eml-entity module to provide detailed information about the logical structure of the dataset. A dataset can be (and often is) composed of a series of data entities (tables) that are linked together by particular integrity constraints. The eml-dataset module, like other modules, may be "referenced" via the <references> tag. This allows a dataset to be described once, and then used as a reference in other locations within the EML document via its ID.

The eml-literature module contains information that describes literature resources. It is intended to provide overview information about the literature citation, including title, abstract, keywords, and contacts. Citation types follow the conventions laid out by EndNote, and there is an attempt to represent a compatible subset of the EndNote citation types. These citation types include: article, book, chapter, edited book, manuscript, report, thesis, conference proceedings, personal communication, map, generic, audio visual, and presentation. The "generic" citation type would be used when one of the other types will not work. The eml-literature module, like other modules, may be "referenced" via the <references> tag. This allows a citation to be described once, and then used as a reference in other locations within the EML document via its ID.

The eml-software module contains general information that describes software resources. This module is intended to fully document software that is needed in order to view a resource (such as a dataset) or to process a dataset. The software module is also imported into the eml-methods module in order to document what software was used to process or perform quality control procedures on a dataset. The eml-software module, like other modules, may be "referenced" via the <references> tag. This allows a software resource to be described once, and then used as a reference in other locations within the EML document via its ID.

The EML Protocol Module is used to define abstract, prescriptive procedures for generating or processing data. Conceptually, a protocol is a standardized method. Eml-protocol resembles eml-methods; however, eml-methods is descriptive (often written in the declarative mood: "I took five subsamples...") whereas eml-protocol is prescriptive (often written in the imperative mood: "Take five subsamples..."). A protocol may have versions, whereas methods (as used in eml-methods) should not.

The following six modules are used to qualify the resources being described in more detail. They are used to describe access control rules, distribution of the metadata and data themselves, parties associated with the resource, the geographic, temporal, and taxonomic extents of the resource, the overall research context of the resource, and detailed methodology used for creating the resource. Some of these modules are imported directly into the top-level resource modules, often in many locations in order to limit the scope of the description. For instance, the eml-coverage module may be used for a particular column of a dataset, rather than the entire dataset as a whole.

The eml-access module describes the level of access that is to be allowed or denied to a resource for a particular user or group of users, and can be described independently for metadata and data. The eml-access module uses a reference to a particular authentication system to determine the set of principals (users or groups) that can be specified in the access rules. The special principal 'public' can be used to indicate that any user or group has access permission, thereby making it easier to specify that anonymous access is allowed. There are two mechanisms for including access control via the eml-access module: The top-level "eml" element may have an optional <access> element that is used to establish the default access control for the entire EML package. If this access element is omitted from the document, then the package submitter should be given full access to the package but all other users should be denied all access. To allow the package to be publicly viewable, the EML author must explicitly include a rule stating so. Barring the existence of a distribution-level <access> element (see below), access to data entities will be controlled by the package-level <access> element in the <eml> element. Exceptions for particular entity-level components of the package can be controlled at a finer grain by using an access description in that entity's physical/distribution tree. When access control rules are specified at this level, they apply only to the data in the parent distribution element, and not to the metadata. Thus, it will control access to the content of the <inline> element, as well as resources that are referenced by the <online/url> and <online/connection> paths. These exceptions to access for particular data resources are applied after the default access rules at the package-level have been applied, so they effectively override the default rules when they overlap. In previous versions of EML access rules for entity-level distribution were contained in <additionalMetadata> sections and referenced via the <describes> tag. Although in theory these could have referenced any node, in application such node-level access control is problematic. Since the most common uses of access control rules were to limit access to specific data entities, the access tree has been placed there explicitly in EML 2.1.0. Access is specified with a choice of child elements, either <allow> or <deny>. Within these rules, values can be assigned for each <principal> using the <permission> element. Users given "read" permission can view the resource; "write" allows changes to the resource excluding changes to the access rules; "changePermission" includes "write" plus the changing of access rules. Users allowed "all" permissions; may do all of the above. Access to data and metadata is affected by the order attribute of the <access> element. It is possible for a deny rule to override an allow rule, and vice versa. In the case where the order attribute is set to "allowFirst", and there are rules similar to the following (with non-critical sections deleted): <deny> <principal>public</principal> <permission>read</permission> </deny> <allow> <principal>uid=alice,o=NASA,dc=ecoinformatics,dc=org</principal> <permission>read</permission> </allow> the principal "uid=alice ..." will be denied access, because it is a member of the special "public" principal, and the deny rule is processed second. For this allow rule to truly allow access to that principal, the order attribute should be set to "denyFirst", and the allow rule will override the deny rule when it is processed second. An example is given below, with non-critical sections deleted: <eml> <access authSystem="ldap://ldap.ecoinformatics.org:389/dc=ecoinformatics,dc=org" order="allowFirst"> <allow> <principal>uid=alice,o=NASA,dc=ecoinformatics,dc=org</principal> <permission>read</permission> <permission>write</permission> <allow> </access> <dataset> ... ... <dataTable id="entity123"> ... <physical> ... <distribution> ... <access id="access123" authSystem="ldap://ldap.ecoinformatics.org:389/dc=ecoinformatics,dc=org" order="allowFirst"> <deny> <principal>uid=alice,o=NASA,dc=ecoinformatics,dc=org</principal> <permission>write</permission> </deny> </access> </distribution> </physical> </dataTable> <dataTable id="entity234"> ... <physical> ... <distribution> ... <access> <references>access123</references> </access> </distribution> </physical> </dataTable> ... </dataset> <eml> In this example, the overall default access is to allow the user=alice (but no one else) to read and write all metadata and data. However, under "entity123" and "entity234", there is an additional rule saying that user=alice does not have write permission. The net effect is that Alice can read and make changes to the metadata, but cannot make changes to the two data entities. In addition, Alice cannot change these access rules; although the submitter can. This example also shows how the eml-access module, like other modules, may be "referenced" via the <references> tag. This allows an access control document to be described once, and then used as a reference in other locations within the EML document via its ID. In summary, access rules can be applied in two places in an eml document. Default access rules are established in the top <access> element for the main eml document (e.g., "/eml/access"). These default rules can be overridden for particular data entities by adding additional <access> elements in the physical/distribution trees of those entities.

The eml-physical module describes the external and internal physical characteristics of a data object as well as the information required for its distribution. Examples of the external physical characteristics of a data object would be the filename, size, compression, encoding methods, and authentication of a file or byte stream. Internal physical characteristics describe the format of the data object being described. Both named binary or otherwise proprietary formats can be cited (e.g., Microsoft Access 2000), or text formats can be precisely described (e.g., ASCII text delimited with commas). For these text formats, it also includes the information needed to parse the data object to extract the entity and its attributes from the data object. Distribution information describes how to retrieve the data object. The retrieval information can be either online (e.g., a URL or other connection information) or offline (e.g., a data object residing on an archival tape). The eml-physical module, like other modules, may be "referenced" via the <references> tag. This allows a physical document to be described once, and then used as a reference in other locations within the EML document via its ID.

The eml-party module describes a responsible party and is typically used to name the creator of a resource or metadata document. A responsible party may be an individual person, an organization or a named position within an organization. The eml-party module contains detailed contact information. It is used throughout the other EML modules where detailed contact information is needed. The eml-party module, like other modules, may be "referenced" via the <references> tag. This allows a party to be described once, and then used as a reference in other locations within the EML document via its ID.

The eml-coverage module contains fields for describing the coverage of a resource in terms of time, space, and taxonomy. These coverages (temporal, spatial, and taxonomic) represent the extent of applicability of the resource in those domains. The Geographic coverage section allows for 2 means of expressing coverage on the surface of the earth: 1) via a set of bounding coordinates that define the North, South, East and West points in a rectangular area, optionally including a bounding altitude, and 2) using a G-Ring polygon definition, where an irregularly shaped area may be defined using a ordered list of latitude/longitude coordinates. A G-Ring may also include an "inner G-Ring" that defines one or more "cut-outs" in the area, i.e. the donut hole concept. The temporal coverage section allows for the definition of either a single date or time, or a range of dates or times. These may be expressed as a calendar date according to the ISO 8601 Date and Time Specification, or by using an alternate time scale, such as the geologic time scale. Currently, EML does not have specific fields to indicate that a data resource may be "ongoing." Two examples are data tables that are planned to be appended in the future, or resources with complex connection definitions (such as to a database) which may return data in real time. It is important that EML be able to handle data from both the "producer" and "consumer" points of view, although currently the temporal coverage modules are designed for the latter. There is no universally acceptable recommendation for describing "ongoing" data within EML. Some groups have chosen to use the <alternateTimeScale> node for the end date, with a value of "ongoing," although this practice is not endorsed by the EML authors. A better solution could be to use very general content for the endDate (such as only the current year) so that the data are accurately described, and searches return datasets as expected. A future version of EML will accommodate such data types with coverage elements specific to their needs. The taxonomic coverage section allows for detailed description of the taxonomic extent of the dataset or resource. The taxonomic classification consists of a recursive set of taxon rank names, their values, and their common names. This construct allows for a taxonomic hierarchy to be built to show the level of identification (e.g. Rank Name = Kingdom, Rank Value = Animalia, Common Name = Animals, and so on down the hierarchy.) The taxonomic coverage module also allows for the definition of the classification system in cases where alternative systems are used. The eml-coverage module, like other modules, may be "referenced" via the <references> tag. This allows the coverage extent to be described once, and then used as a reference in other locations within the EML document via its ID.

The eml-project module describes the research context in which the dataset was created, including descriptions of over-all motivations and goals, funding, personnel, description of the study area etc. This is also the module to describe the design of the project: the scientific questions being asked, the architecture of the design, etc. This module is used to place the dataset that is being documented into its larger research context. The eml-project module, like other modules, may be "referenced" via the <references> tag. This allows a research project to be described once, and then used as a reference in other locations within the EML document via its ID.

The eml-methods module describes the methods followed in the creation of the dataset, including description of field, laboratory and processing steps, sampling methods and units, quality control procedures. The eml-methods module is used to describe the actual procedures that are used in the creation or the subsequent processing of a dataset. Likewise, eml-methods is used to describe processes that have been used to define / improve the quality of a data file, or to identify potential problems with the data file. Note that the eml-protocol module is intended to be used to document a prescribed procedure, whereas the eml-method module is used to describe procedures that were actually performed. The distinction is that the use of the term "protocol" is used in the "prescriptive" sense, and the term "method" is used in the "descriptive" sense. This distinction allows managers to build a protocol library of well-known, established protocols (procedures), but also document what procedure was truly performed in relation to the established protocol. The method may have diverged from the protocol purposefully, or perhaps incidentally, but the procedural lineage is still preserved and understandable.

The following three modules are used to document the logical layout of a dataset. Many datasets are comprised of multiple entities (e.g. a series of tabular data files, or a set of GIS features, or a number of tables in a relational database). Each entity within a dataset may contain one or more attributes (e.g. multiple columns in a data file, multiple attributes of a GIS feature, or multiple columns of a database table). Lastly, there may be both simple or complex relationships among the entities within a dataset. The relationships, or the constraints that are to be enforced in the dataset, are described using the eml-constraint module. All entities share a common set of information (described using eml-entity), but some discipline specific entities have characteristics that are unique to that entity type. Therefore, the eml-entity module is extended for each of these types (dataTable, spatialRaster, spatialVector, etc...) which are described in the next section.

The eml-entity module defines the logical characteristics of each entity in the dataset. Entities are usually tables of data (eml-dataTable). Data tables may be ascii text files, relational database tables, spreadsheets or other type of tabular data with a fixed logical structure. Related to data tables are views (eml-view) and stored procedures (eml-storedProcedure). Views and stored procedures are produced by an RDBMS or related system. Other types of data such as: raster (eml-spatialRaster), vector (eml-spatialVector) or spatialReference image data are also data entities. An otherEntity element would be used to describe types of entities that are not described by any other entity type. The eml-entity module, like other modules, may be "referenced" via the <references> tag. This allows an entity document to be described once, and then used as a reference in other locations within the EML document via its ID.

The eml-attribute module describes all attributes (variables) in a data entity: dataTable, spatialRaster, spatialVector, storedProcedure, view or otherEntity. The description includes the name and definition of each attribute, its domain, definitions of coded values, and other pertinent information. Two structures exist in this module: 1. attribute is used to define a single attribute; 2. attributeList is used to define a list of attributes that go together in some logical way. The eml-attribute module, like other modules, may be "referenced" via the <references> tag. This allows an attribute document to be described once, and then used as a reference in other locations within the EML document via its ID.

The concept of "unit" represents one of the most fundamental categories of metadata. The classic example of data entropy is the case in which a reported numeric value loses meaning due to lack of associated units. Much of Ecology is driven by measurement, and most measurements are inherently comparative. Good data description requires a representation of the basis for comparison, i.e., the unit. In modeling the attribute element, the authors of EML drew inspiration from the NIST Reference on Constants, Units, and Uncertainty. This document defines a unit as "a particular physical quantity, defined and adopted by convention, with which other particular quantities of the same kind are compared to express their value." The authors of the EML 2.0 specification (hereafter "the authors") decided to make the unit element required, wherever possible. Units may also be one of the most problematic categories of metadata. For instance, there are many candidate attributes that clearly have no units, such as named places and letter grades. There are other candidate attributes for which units are difficult to identify, despite some suspicion that they should exist (e.g. pH, dates, times). In still other cases, units may be meaningful, but apparently absent due to dimensional analysis (e.g. grams of carbon per gram of soil). The relationship between units and dimensions likewise is not completely clear. The authors decided to sharpen the model of attribute by nesting unit under measurementScale. Measurement Scale is a data typology, borrowed from Statistics, that was introduced in the 1940's. Under the adopted model, attributes are classified as nominal, ordinal, interval, and ratio. Though widely criticized, this classification is well-known and provides at least first-order utility in EML. For example, nesting unit under measurementScale allows EML to prevent its meaningless inclusion for categorical data -- an approach judged superior to making unit universally required or universally optional. The sharpening of the attribute model allowed the elimination of the unit type "undefined" from the standard unit dictionary (see eml-unitDictionary.xml). It seemed self-defeating to require the unit element exactly where appropriate, yet still allow its content to be undefined. An attribute that requires a unit definition is malformed until one is provided. The unit type "dimensionless" is preserved, however. In EML 2.0, it is synonymous with "unitless" and represents the case in which units cannot be associated with an attribute for some reason, despite the proper classification of that attribute as interval or ratio. Dimensionless may itself be an anomaly arising from the limitations of the adopted measurement scale typology. Closely related to the concept of unit is the concept of attribute domain. The authors decided that a well-formed description of an attribute must include some indication of the set of possible values for that attribute. The set of possible values is useful, perhaps necessary, for interpreting any particular observed value. While universally required, attribute domain has different forms, depending on the associated measurement scale. The element storageType has an obvious relationship to domain. It gives some indication of the range of possible values of an attribute, and also gives some (potentially critical) operability information about the way the attribute is represented or construed in the local storage system. The storageType element seems to fall in a gray area between the logical and physical aspects of stored data. Neither comfortable with eliminating it nor with making it required, the authors left it available but optional under attribute. In addition, it is repeatable so that different storage types can be provided for various systems (e.g., different databases might use different types for columns, even though the domain of the attribute is the same regardless of which database is used). Attributes representing dates, times, or combinations thereof (hereafter "dateTime") were the most difficult to model in EML. Is dateTime of type interval or ordinal? Does it have units or not? Strong cases can be made on each side of the issue. The confusion may reflect the limitations of the measurement scale typology. The final resolution of the dateTime model is probably somewhat arbitrary. There was clearly a need, however, to allow for the interoperability of dateTime formats. EML 2.0 tries to provide an unambiguous mechanism for describing the format of dateTime values by providing a separate category for date and time values. This "dateTime" measurement scale allows users to explicitly label attributes that contain Gregorian date and time values, and allows them to provide the information needed to parse these values into their appropriate components (e.g., days, months, years)./

The eml-constraint schema defines the integrity constraints between entities (e.g., data tables) as they would be maintained in a relational management system. These constraints include primary key constraints, foreign key constraints, unique key constraints, check constraints, and not null constraints, among potential others.

The following six modules are used to describe a number of common types of entities found in datasets. Each entity type uses the eml-entity module elements as it's base set of elements, but then extends the base with entity-specific elements. Note that the eml-spatialReference module is not an entity type, but is rather a common set of elements used to describe spatial reference systems in both eml-spatialRaster and eml-spatialVector. It is described here in relation to those two modules.

The eml-dataTable module is used to describe the logical characteristics of each tabular set of information in a dataset. A series of comma-separated text files may be considered a dataset, and each file would subsequently be considered a dataTable entity within the dataset. Since the eml-dataTable module extends the eml-entity module, it uses all of the common entity elements to describe the table, along with a few elements specific to just data table entities. The eml-dataTable module allows for the description of each attribute (column/field/variable) within the data table through the use of the eml-attribute module. Likewise, there are fields used to describe the physical distribution of the data table, its overall coverage, the methodology used in creating the data, and other logical structure information such as its orientation, case sensitivity, etc.

The eml-spatialRaster module allows for the description of entities composed of rectangular grids of data values that are usually georeferenced to a portion of the earth's surface. Specific attributes of a spatial raster can be documented here including the spatial organization of the raster cells, the cell data values, and if derived via imaging sensors, characteristics about the image and its individual bands.

The eml-spatialVector module allows for the description of spatial objects in a GIS system that are not defined in a regularly gridded pattern. These geometries include points and vectors and the relationships among them. Specific attributes of a spatial vector can be documented here including the vector's geometry type, count and topology level.

This module defines both projected and unprojected coordinate systems for referencing the spatial coordinates of a dataset to the earth. The schema is based on that used by Environmental Systems Research Inc (ESRI) for its .prj file format. EML provides a library of pre-defined coordinate systems that may be referred to by name in the horizCoordSysName element. A custom projection may be defined using this schema for any projection that does not appear in this dictionary.

The storedProcedure module is meant to capture information on procedures that produce data output in the form of a data table. In an RDBMS one can code complex queries and transactions into stored procedures and then invoke them directly from front-end applications. It allows the optional description of any parameters that are expected to be passed to the procedure when it is called.

The eml-view module describes a view from a database management system. A view is a query statement that is stored as a database object and executed each time the view is called.

The following modules are used to highlight the information being documented in each of the above modules where prose may be needed to convey the critical metadata. The eml-text module provides a number of text-based constructs to enhance a document (including sections, paragraphs, lists, subscript, superscript, emphasis, etc.)

The eml-text module is a wrapper container that allows general text descriptions to be used within the various modules of eml. It can include either structured or unstructured text blocks. It isn't really appropriate to use this module outside of the context of a parent module, because the parent module determines the appropriate context to which this text description applies. The eml-text module allows one to provide structure to a text description in order to convey concepts such as sections (paragraphs), hierarchy (ordered and unordered lists), emphasis (bold, superscript, subscript) etc. The structured elements are a subset of DocBook so the predefined DocBook stylesheets can be used to style EML fields that implement this module.

The multiple modules in EML all depend on each other in complex ways. To easily see these dependencies see the EML Dependency Chart.

EML supports internationalization using the i18nNonEmptyStringType. Fields defined as this type include: Title Keyword Contact information (e.g. names, organizations, addresses) TextType fields also support language translations. These fields include: Abstract Methods Protocol Core metadata should be provided in English. The core elements can be augmented with translations in a native language. Detailed metadata can be provided in the native language as declared using the xml:lang attribute. Authors can opt to include English translations of this detailed metadata as they see fit. The following example metadata document is provided primarily in Portuguese but includes English translations of core metadata fields. The xml:lang="pt_BR" attribute at the root of the EML document indicates that, unless otherwise specified, the content of the document is supplied in Portuguese (Brazil). The xml:lang="en_US" attributes on child elements denote that the content of that element is provided in English. Core metadata (i.e. title) is provided in English, supplemented with a Portuguese translation using the value tag with an xml:lang attribute. Note that child elements can override the root language declaration of the document as well as the language declaration of their containing elements. The abstract element is primarily given in Portuguese (as inherited from the root language declaration), with an English translation. Many EML fields are repeatable (i.e. keyword) so that multiple values can be provided for the same concept. Translations for these fields should be included as nested value tags to indicate that they are equivalent concepts expressed in different languages rather than entirely different concepts. <?xml version="1.0"?> <eml:eml packageId="eml.1.1" system="knb" xml:lang="pt_BR" xmlns:eml="eml://ecoinformatics.org/eml-2.1.1" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="eml://ecoinformatics.org/eml-2.1.1 eml.xsd"> <dataset id="ds.1"> <!-- English title with Portuguese translation --> <title xml:lang=""en_US"> Sample Dataset Description <value xml:lang="pt_BR">Exemplo Descrição Dataset</value> </title> ... <!-- Portuguese abstract with English translation --> <abstract> <para> Neste exemplo, a tradução em Inglês é secundário <value xml:lang="en_US">In this example, the English translation is secondary</value> <para> </abstract> ... <!-- two keywords, each with an equivalent translation --> <keywordSet> <keyword keywordType="theme"> árvore <value xml:lang="en_US">tree</value> <keyword> <keyword keywordType="theme"> água <value xml:lang="en_US">water</value> <keyword> </keywordSet> ... </dataset> </eml:eml>

This section explains the rules of EML. There are some rules that cannot be written directly into the XML Schemas nor enforced by an XML parser. These are guidelines that every EML package must follow in order for it to be considered EML compliant.

Each EML module, with the exception of "eml" itself, has a top level choice between the structured content of that modules or a "references" field. This enables the reuse of content previously defined elsewhere in the document. Methods for defining and referencing content are described in the next section

EML allows the reuse of previously defined structured content (DOM sub-trees) through the use of key/keyRef type references. In order for an EML package to remain cohesive and to allow for the cross platform compatibility of packages, the following rules with respect to packaging must be followed. An ID is required on the eml root element. IDs are optional on all other elements. If an ID is not provided, that content must be interpreted as representing a distinct object. If an ID is provided for content then that content is distinct from all other content except for that content that references its ID. If a user wants to reuse content to indicate the repetition of an object, a reference must be used. Two identical ids with the same system attribute cannot exist in a single document. "Document" scope is defined as identifiers unique only to a single instance document (if a document does not have a system attribute or if scope is set to 'document' then all IDs are defined as distinct content). "System" scope is defined as identifiers unique to an entire data management system (if two documents share a system string, then any IDs in those two documents that are identical refer to the same object). If an element references another element, it must not have an ID itself. The system attribute must have the same value in both the target and referencing elements or it must be absent in both. All EML packages must have the 'eml' module as the root. The system and scope attribute are always optional except for at the 'eml' module where the scope attribute is fixed as 'system'. The scope attribute defaults to 'document' for all other modules.

Because some of these rules cannot be enforced in XML-Schema, we have written a parser which checks the validity of the references and IDs used in your document. This parser is included with the 2.1.0 release of EML. To run the parser, you must have Java 1.3.1 or higher. To execute it change into the lib directory of the release and run the 'runEMLParser' script passing your EML instance file as a parameter. There is also an online version of this parser which is publicly accessible. The online parser will both validate your XML document against the schema as well as check the integrity of your references.

<?xml version="1.0"?> <eml:eml packageId="eml.1.1" system="knb" xmlns:eml="eml://ecoinformatics.org/eml-2.1.1" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="eml://ecoinformatics.org/eml-2.1.1 eml.xsd"> <dataset id="ds.1"> <title>Sample Dataset Description</title> <!-- the two creators have the same id. this should be an error--> <creator id="23445" scope="document"> <individualName> <surName>Smith</surName> </individualName> </creator> <creator id="23445" scope="document"> <individualName> <surName>Myer</surName> </individualName> </creator> ... </dataset> </eml:eml>

This instance document is invalid because both creator elements have the same id. No two elements can have the same string as an id.

<?xml version="1.0"?> <eml:eml packageId="eml.1.1" system="knb" xmlns:eml="eml://ecoinformatics.org/eml-2.1.1" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="eml://ecoinformatics.org/eml-2.1.1 eml.xsd"> <dataset id="ds.1"> <title>Sample Dataset Description</title> <creator id="23445" scope="document"> <individualName> <surName>Smith</surName> </individualName> </creator> <creator id="23446" scope="document"> <individualName> <surName>Myer</surName> </individualName> </creator> ... <contact> <references>23447</references> </contact> </dataset> </eml:eml>

This instance document is invalid because the contact element references an id that does not exist. Any referenced id must exist.

<?xml version="1.0"?> <eml:eml packageId="eml.1.1" system="knb" xmlns:eml="eml://ecoinformatics.org/eml-2.1.1" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="eml://ecoinformatics.org/eml-2.1.1 eml.xsd"> <dataset id="ds.1"> <title>Sample Dataset Description</title> <creator id="23445" scope="document"> <individualName> <surName>Smith</surName> </individualName> </creator> <creator id="23446" scope="document"> <individualName> <surName>Meyer</surName> </individualName> </creator> ... <contact id="522"> <references>23445</references> </contact> </dataset> </eml:eml>

This instance document is invalid because the contact element both references another element and has an id itself. If an element references another element, it may not have an id. This prevents circular references.

<?xml version="1.0"?> <eml:eml packageId="eml.1.1" system="knb" xmlns:eml="eml://ecoinformatics.org/eml-2.1.1" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="eml://ecoinformatics.org/eml-2.1.1 eml.xsd"> <dataset id="ds.1"> <title>Sample Dataset Description</title> <creator id="23445" scope="document"> <individualName> <surName>Smith</surName> </individualName> </creator> <creator id="23446" scope="document"> <individualName> <surName>Smith</surName> </individualName> </creator> ... <contact> <references>23446</references> </contact> <contact> <references>23445</references> </contact> </dataset> </eml:eml>

This instance document is valid. Each contact is referencing one of the creators above and all the ids are unique.

Normative technical docs for eml

Normative technical docs for eml-access

Normative technical docs for eml-attribute

Normative technical docs for eml-constraint

Normative technical docs for eml-coverage

Normative technical docs for eml-dataset

Normative technical docs for eml-dataTable

Normative technical docs for eml-entity

Normative technical docs for eml-literature

Normative technical docs for eml-methods

Normative technical docs for eml-party

Normative technical docs for eml-physical

Normative technical docs for eml-project

Normative technical docs for eml-protocol

Normative technical docs for eml-resource

Normative technical docs for eml-software

Normative technical docs for eml-spatialRaster

Normative technical docs for eml-spatialReference

Normative technical docs for eml-spatialVector

Normative technical docs for eml-storedProcedure

Normative technical docs for eml-text

Normative technical docs for eml-unitTypeDefinitions

Normative technical docs for eml-view

abstract-eml-project abstract-eml-resource access-eml access-eml-access access-eml-physical accuracy-eml-attribute accuracyReport-eml-spatialRaster accuracyReport-eml-spatialVector action-eml-software additionalInfo-eml-entity additionalInfo-eml-resource additionalMetadata-eml address-eml-party administrativeArea-eml-party allow-eml-access alternateIdentifier-eml-entity alternateIdentifier-eml-resource alternativeTimeScale-eml-coverage altitudeDatumName-eml-spatialReference altitudeDistanceUnits-eml-spatialReference altitudeEncodingMethod-eml-spatialReference altitudeMaximum-eml-coverage altitudeMinimum-eml-coverage altitudeResolution-eml-spatialReference altitudeSysDef-eml-spatialReference altitudeUnits-eml-coverage article-eml-literature associatedParty-eml-resource attribute-eml-attribute attribute-eml-attribute attributeAccuracyExplanation-eml-attribute attributeAccuracyReport-eml-attribute attributeAccuracyValue-eml-attribute attributeDefinition-eml-attribute attributeLabel-eml-attribute attributeList-eml-attribute attributeList-eml-dataTable attributeList-eml-entity attributeList-eml-spatialRaster attributeList-eml-spatialVector attributeList-eml-storedProcedure attributeList-eml-view attributeName-eml-attribute attributeOrientation-eml-physical attributeReference-eml-constraint attributeReference-eml-constraint attributeReference-eml-constraint attributeReference-eml-constraint attributeReference-eml-constraint audioVisual-eml-literature authentication-eml-physical bandDescription-eml-spatialRaster bandgapbytes-eml-physical bandrowbytes-eml-physical beginDate-eml-coverage bilinearFit-eml-spatialRaster binaryRasterFormat-eml-physical book-eml-literature bookTitle-eml-literature boundingAltitudes-eml-coverage boundingCoordinates-eml-coverage bounds-eml-attribute bounds-eml-attribute byteorder-eml-physical calendarDate-eml-coverage cameraCalibrationInformationAvailability-eml-spatialRaster cardinality-eml-constraint caseSensitive-eml-dataTable cellGeometry-eml-spatialRaster cellSizeXDirection-eml-spatialRaster cellSizeYDirection-eml-spatialRaster changeDate-eml-dataset changeHistory-eml-dataset changeScope-eml-dataset chapter-eml-literature chapterNumber-eml-literature characterEncoding-eml-physical checkCondition-eml-constraint checkConstraint-eml-constraint checksum-eml-software childOccurences-eml-constraint citation-eml citation-eml-attribute citation-eml-literature citation-eml-methods citation-eml-methods citation-eml-physical citation-eml-project citation-eml-project citation-eml-project citetitle-eml-text city-eml-party classificationSystem-eml-coverage classificationSystemCitation-eml-coverage classificationSystemModifications-eml-coverage cloudCoverPercentage-eml-spatialRaster code-eml-attribute code-eml-attribute codeDefinition-eml-attribute codeExplanation-eml-attribute codesetName-eml-attribute codesetURL-eml-attribute collapseDelimiters-eml-physical collapseDelimiters-eml-physical column-eml-spatialRaster columns-eml-spatialRaster comment-eml-dataset commonName-eml-coverage communicationType-eml-literature complex-eml-physical compressionGenerationQuality-eml-spatialRaster compressionMethod-eml-physical conferenceDate-eml-literature conferenceDate-eml-literature conferenceLocation-eml-literature conferenceLocation-eml-literature conferenceName-eml-literature conferenceName-eml-literature conferenceProceedings-eml-literature connection-eml-physical connection-eml-resource connectionDefinition-eml-resource connectionDefinition-eml-resource constraint-eml-dataTable constraint-eml-entity constraint-eml-spatialRaster constraint-eml-spatialVector constraint-eml-storedProcedure constraint-eml-view constraintDescription-eml-constraint constraintName-eml-constraint contact-eml-dataset contact-eml-literature controlPoint-eml-spatialRaster corner-eml-spatialRaster cornerPoint-eml-spatialRaster country-eml-party coverage-eml-attribute coverage-eml-entity coverage-eml-methods coverage-eml-methods coverage-eml-project coverage-eml-resource creator-eml-resource customUnit-eml-attribute dataFormat-eml-physical dataset-eml dataset-eml-dataset datasetGPolygon-eml-coverage datasetGPolygonExclusionGRing-eml-coverage datasetGPolygonOuterGRing-eml-coverage dataSource-eml-methods dataTable-eml-dataset dataTable-eml-dataTable dateTime-eml-attribute dateTimeDomain-eml-attribute dateTimePrecision-eml-attribute datum-eml-spatialReference defaultValue-eml-resource definition-eml-attribute definition-eml-attribute definition-eml-resource definitionAttributeReference-eml-attribute degree-eml-literature deliveryPoint-eml-party deny-eml-access dependency-eml-software depthDatumName-eml-spatialReference depthDistanceUnits-eml-spatialReference depthEncodingMethod-eml-spatialReference depthResolution-eml-spatialReference depthSysDef-eml-spatialReference describes-eml description-eml-dataset description-eml-methods description-eml-methods description-eml-project description-eml-resource descriptor-eml-project descriptorValue-eml-project designDescription-eml-project diskUsage-eml-software distribution-eml-physical distribution-eml-resource distribution-eml-software domainDescription-eml-storedProcedure eastBoundingCoordinate-eml-coverage editedBook-eml-literature edition-eml-literature edition-eml-literature edition-eml-literature editor-eml-literature electronicMailAddress-eml-party eml-eml emphasis-eml-text encodingMethod-eml-physical endDate-eml-coverage entityCodeList-eml-attribute entityDescription-eml-entity entityName-eml-entity entityReference-eml-attribute entityReference-eml-constraint entityType-eml-entity enumeratedDomain-eml-attribute externalCodeSet-eml-attribute externallyDefinedFormat-eml-physical fieldDelimiter-eml-physical fieldDelimiter-eml-physical fieldStartColumn-eml-physical fieldWidth-eml-physical filmDistortionInformationAvailability-eml-spatialRaster foreignKey-eml-constraint formatName-eml-physical formatString-eml-attribute formatVersion-eml-physical funding-eml-project generalTaxonomicCoverage-eml-coverage generic-eml-literature geogCoordSys-eml-spatialReference geogCoordSys-eml-spatialReference geographicCoverage-eml-coverage geographicCoverage-eml-literature geographicDescription-eml-coverage geometricObjectCount-eml-spatialVector geometry-eml-spatialVector georeferenceInfo-eml-spatialRaster givenName-eml-party gRing-eml-coverage gRing-eml-coverage gRingLatitude-eml-coverage gRingLongitude-eml-coverage gRingPoint-eml-coverage gRingPoint-eml-coverage highWavelength-eml-spatialRaster horizCoordSysDef-eml-spatialReference horizCoordSysDef-eml-spatialReference horizCoordSysName-eml-spatialReference horizontalAccuracy-eml-spatialRaster horizontalAccuracy-eml-spatialVector identificationReference-eml-coverage identifierName-eml-coverage illuminationAzimuthAngle-eml-spatialRaster illuminationElevationAngle-eml-spatialRaster imageDescription-eml-spatialRaster imageOrientationAngle-eml-spatialRaster imageQualityCode-eml-spatialRaster imagingCondition-eml-spatialRaster implementation-eml-software individualName-eml-party inline-eml-physical inline-eml-resource institution-eml-literature institution-eml-literature instrumentation-eml-methods intellectualRights-eml-resource interval-eml-attribute issue-eml-literature itemizedlist-eml-text itemizedlist-eml-text joinCondition-eml-constraint journal-eml-literature key-eml-constraint key-eml-constraint key-eml-constraint key-eml-constraint keyword-eml-resource keywordSet-eml-resource keywordThesaurus-eml-resource language-eml-resource language-eml-software layout-eml-physical lensDistortionInformationAvailability-eml-spatialRaster license-eml-software licenseURL-eml-software lineNumber-eml-physical lineNumber-eml-physical listitem-eml-text literalCharacter-eml-physical literalCharacter-eml-physical literalLayout-eml-text lowWaveLength-eml-spatialRaster machineProcessor-eml-software maintenance-eml-dataset maintenanceUpdateFrequency-eml-dataset manuscript-eml-literature map-eml-literature maximum-eml-attribute maximum-eml-attribute maxRecordLength-eml-physical measurementScale-eml-attribute mediumDensity-eml-resource mediumDensityUnits-eml-resource mediumFormat-eml-resource mediumName-eml-resource mediumNote-eml-resource mediumVolume-eml-resource metadata-eml metadataProvider-eml-resource methods-eml-attribute methods-eml-dataset methods-eml-entity methods-eml-methods methodStep-eml-methods minimum-eml-attribute minimum-eml-attribute missingValueCode-eml-attribute multiBand-eml-physical name-eml-resource name-eml-resource name-eml-storedProcedure nbands-eml-physical nbits-eml-physical nominal-eml-attribute nonNumericDomain-eml-attribute nonNumericDomain-eml-attribute northBoundingCoordinate-eml-coverage notNullConstraint-eml-constraint numberOfBands-eml-spatialRaster numberOfRecords-eml-dataTable numberOfVolumes-eml-literature numberOfVolumes-eml-literature numberType-eml-attribute numericDomain-eml-attribute numericDomain-eml-attribute numFooterLines-eml-physical numHeaderLines-eml-physical numPhysicalLinesPerRecord-eml-physical objectName-eml-physical offline-eml-physical offline-eml-resource offset-eml-spatialRaster oldValue-eml-dataset online-eml-physical online-eml-resource onlineDescription-eml-physical onlineDescription-eml-resource onlineUrl-eml-party operatingSystem-eml-software orderAttributeReference-eml-attribute orderedlist-eml-text orderedlist-eml-text ordinal-eml-attribute organizationName-eml-party originalPublication-eml-literature originator-eml-coverage otherEntity-eml-dataset otherEntity-eml-entity pageRange-eml-literature pageRange-eml-literature para-eml-text para-eml-text para-eml-text parameter-eml-resource parameter-eml-spatialReference parameter-eml-storedProcedure parameterDefinition-eml-resource parentOccurences-eml-constraint party-eml-party pattern-eml-attribute peakResponse-eml-spatialRaster performer-eml-literature permission-eml-access personalCommunication-eml-literature personnel-eml-project phone-eml-party physical-eml-entity physical-eml-physical physicalLineDelimiter-eml-physical pointInPixel-eml-spatialRaster pointInPixel-eml-spatialRaster positionName-eml-party postalCode-eml-party precision-eml-attribute precision-eml-attribute preProcessingTypeCode-eml-spatialRaster presentation-eml-literature primaryKey-eml-constraint primeMeridian-eml-spatialReference principal-eml-access proceduralStep-eml-protocol programmingLanguage-eml-software projCoordSys-eml-spatialReference project-eml-dataset project-eml-software projection-eml-spatialReference projectionList-eml-spatialReference protocol-eml protocol-eml-methods protocol-eml-protocol pubDate-eml-resource publicationPlace-eml-literature publicationPlace-eml-literature publicationPlace-eml-literature publicationPlace-eml-literature publicationPlace-eml-literature publicationPlace-eml-literature publisher-eml-dataset publisher-eml-literature publisher-eml-literature publisher-eml-literature publisher-eml-literature publisher-eml-literature publisher-eml-literature publisher-eml-literature pubPlace-eml-dataset purpose-eml-dataset qualityControl-eml-methods quantitativeAccuracyMethod-eml-spatialRaster quantitativeAccuracyMethod-eml-spatialVector quantitativeAccuracyReport-eml-spatialRaster quantitativeAccuracyReport-eml-spatialVector quantitativeAccuracyValue-eml-spatialRaster quantitativeAccuracyValue-eml-spatialVector quantitativeAttributeAccuracyAssessment-eml-attribute queryStatement-eml-view quoteCharacter-eml-physical quoteCharacter-eml-physical radiometricDataAvailability-eml-spatialRaster rangeOfDates-eml-coverage rasterOrigin-eml-spatialRaster ratio-eml-attribute recipient-eml-literature recordDelimiter-eml-physical referencedEntityId-eml-methods referencedKey-eml-constraint references-eml-resource referenceType-eml-literature relatedProject-eml-project relationshipType-eml-constraint repeats-eml-storedProcedure report-eml-literature reportNumber-eml-literature repository-eml-coverage reprintEdition-eml-literature required-eml-storedProcedure researchProject-eml-project reviewedItem-eml-literature role-eml-project role-eml-resource row-eml-spatialRaster rowColumnOrientation-eml-physical rows-eml-spatialRaster runtimeMemoryUsage-eml-software salutation-eml-party sampling-eml-methods samplingDescription-eml-methods scale-eml-literature scaleFactor-eml-spatialRaster schemeName-eml-resource section-eml-text section-eml-text sequenceIdentifier-eml-spatialRaster series-eml-resource shortName-eml-resource simpleDelimited-eml-physical singleDateTime-eml-coverage size-eml-physical size-eml-software skipbytes-eml-physical software-eml software-eml-methods software-eml-software source-eml-attribute source-eml-attribute southBoundingCoordinate-eml-coverage spatialRaster-eml-dataset spatialRaster-eml-spatialRaster spatialReference-eml-spatialRaster spatialReference-eml-spatialReference spatialReference-eml-spatialVector spatialSamplingUnits-eml-methods spatialVector-eml-dataset spatialVector-eml-spatialVector specimen-eml-coverage spheroid-eml-spatialReference standardUnit-eml-attribute storageType-eml-attribute storedProcedure-eml-dataset storedProcedure-eml-storedProcedure studyAreaDescription-eml-project studyExtent-eml-methods subscript-eml-text subscript-eml-text subStep-eml-methods superscript-eml-text superscript-eml-text surName-eml-party taxonomicClassification-eml-coverage taxonomicClassification-eml-coverage taxonomicCompleteness-eml-coverage taxonomicCoverage-eml-coverage taxonomicProcedures-eml-coverage taxonomicSystem-eml-coverage taxonRankName-eml-coverage taxonRankValue-eml-coverage temporalCoverage-eml-coverage text-eml-text textDelimited-eml-physical textDomain-eml-attribute textFixed-eml-physical textFormat-eml-physical thesis-eml-literature time-eml-coverage timeScaleAgeEstimate-eml-coverage timeScaleAgeExplanation-eml-coverage timeScaleAgeUncertainty-eml-coverage timeScaleCitation-eml-coverage timeScaleName-eml-coverage title-eml-project title-eml-resource title-eml-text toneGradation-eml-spatialRaster topologyLevel-eml-spatialVector totalFigures-eml-literature totalFigures-eml-literature totalPages-eml-literature totalPages-eml-literature totalPages-eml-literature totalPages-eml-literature totalPages-eml-literature totalrowbytes-eml-physical totalTables-eml-literature totalTables-eml-literature triangulationIndicator-eml-spatialRaster ulink-eml-text uniqueKey-eml-constraint unit-eml-attribute unit-eml-attribute unit-eml-spatialReference unit-eml-spatialReference url-eml-physical url-eml-resource userId-eml-party value-eml-resource value-eml-resource value-eml-text value-eml-text value-eml-text value-eml-text valueAttributeReference-eml-attribute version-eml-software vertCoordSys-eml-spatialReference verticalAccuracy-eml-spatialRaster verticalAccuracy-eml-spatialVector verticals-eml-spatialRaster view-eml-dataset view-eml-view virtualMachine-eml-software volume-eml-literature volume-eml-literature volume-eml-literature vouchers-eml-coverage waveLengthUnits-eml-spatialRaster westBoundingCoordinate-eml-coverage xCoordinate-eml-spatialRaster xCoordinate-eml-spatialRaster xIntercept-eml-spatialRaster xSlope-eml-spatialRaster yCoordinate-eml-spatialRaster yCoordinate-eml-spatialRaster yIntercept-eml-spatialRaster ySlope-eml-spatialRaster