NHD Feature Templates
Relevance and Use of NHD Feature Catalog
Based on 1:24,000-scale USGS Topographic Maps
About Standards for Hydrography
Standards for the depiction of hydrography on maps have been an integral part of USGS mapping for a century. With some modifications, the hydrographic mapping standards were used as the foundation for the NHD. These standards will give way to the development of “best practices” by stewards and NHD users. Despite the transition, the information in the catalog remains a valuable resource for users, because it documents decades of thinking about hydrographic data. Similarly, sections in the individual feature templates on ‘Source Interpretation Guidelines’ can be used by data stewards to document the revision techniques they use.
In the mid 1990’s, USGS developed standards, based on Feature Templates, to describe the content revision requirements for standard USGS topographic maps. Standards were documented for all of the data themes on topographic maps, including transportation, hypsography, manmade structures, and others. The feature templates were based on the almost 100 years of topographic instructions for USGS maps. The original maps were created based on field work to identify features on the ground and photogrammetry to delineate the features. Once the maps were produced, most of the updating was based on monoscopic, black and white orthoimagery. Revision content requirements were based on features that could be updated from these black and white, mono sources, with only limited field work done.
The templates were developed using rigorous rules for describing content. A feature-based classification methodology was developed (see following section - Methodology for Identifying Features) and features and attributes were defined using this methodology. Rules for feature delineation, representation and collection (capture conditions) were defined as well as rules for revising features using a variety of sources.
The document, Template Development and Use, explains each section in the feature templates. This document was modified slightly from the 4/96 version and is included as an introduction to the NHD standards. As described in Template Development and Use:
“The capture conditions reflect the NMD Policy of cartographic collection, which restricts data content and position based on graphic limitations of scale and legibility. Only content that can be displayed on the printed map is collected and all offsets in position necessary to accommodate symbolization are performed at the time of data collection. This policy and the feature templates will change as user requirements for geographic content and position become known and as technology and resources are developed to
support the implementation of product generation rules for content generalization and symbol conflict resolution.”
The feature templates for hydrography still contain much valuable information: feature and attribute definitions, rules for valuing the attributes, and the rules that describe what features were shown on USGS topographic maps, and subsequently, what features were collected in the initial population of the high-resolution NHD. These features are generally described as “map-worthy” and were defined as those suitable for a “general-purpose map;” they could be shown on a 1:24,000-scale map without sacrificing legibility. Another factor was the cost of data collection and the reliability of that collection. Features for which photointerpretation was not possible, and field-checking costs were prohibitive, were not generally included. However, as local data stewards assume responsibility for update and revision, the 1:24,000-scale map is not necessarily the benchmark for data collection. The NHD is still defined as a “general-purpose” product and some local details may not be suitable for the national holdings. The legibility concerns for map scales are not limitations for a digital product, but the level of detail may matter a great deal to data modelers and others. The cost of maintenance still applies, however. Features may be relatively easy to collect initially, but verifying the status of that feature can become costly.
The standards have been modified slightly to make them available as the foundation for the NHD. The attributes that apply to the feature classes that became part of the NHD implementation in ArcGIS are included in the attribute/attribute value list, but they are “grayed out” (ComID, FDate, Resolution, AreaSqKm, FType and Fcode). Only the attributes that were defined for the individual feature characteristics are shown in regular text color. Name and Elevation are also “grayed out” if they are NOT collected for a particular feature type. These standards will likely remain as is, with no further maintenance. In keeping with the dynamic nature of the NHD, the standards will be a starting point for the development of “best practices” guidance that the user community can help establish. However, the information in the catalog is still a valuable resource for users as a way to understand the content of the NHD, and the sections in the templates on ‘Source Interpretation Guidelines’ can be used to document the revision techniques used by the data stewards.
Using Feature Templates with the NHD
The information in the feature catalog should still be understood and described as a part of any new data collection or revision. The feature template for lake/pond is a good example of how the catalog can be useful for thinking about features in the NHD. This template has many rules for determining the edge of the lake/pond, and this information is not based on map or scale considerations.
Example of a Feature Delineation
These rules may seem stilted (they were developed for an automatic, rules-based decision processes), but they do reflect information that a data steward needs to consider.
It is not easy--but it is critical--to delineate impounded lake/ponds. USGS “topographic instructions” probably have more pages of descriptions for this than for almost any other feature. On the 1:24,000-scale topographic maps, USGS used the “normal pool” or the “spillway elevation” depending on how stable the water elevation was. The agency that managed the impounded lake/pond was contacted for this information. Because the maps were originally created using stereo photographs, the normal pool or spillway elevation was delineated as the edge of the lake/pond, regardless of the water level in the photos. During revision by data stewards, if elevation data is not available, and the impounded lake/pond already exists in the NHD, we recommend data stewards NOT revise the delineation, even if the water level looks different on an orthoimage. Impounded lake/ponds can be used for flood control, water supply, irrigation, and power generation, and water levels can vary several feet a day. It is more useful to delineate the lake/pond at a stage that can be maintained over time. Otherwise, the lake/pond would have to be revised every time a new photograph is taken. Various plans are underway to link the National Inventory of Dams to the NHD. Once that is complete, information on dams and impoundments will be readily available.
An example of a different delineation for lake/pond comes from Minnesota. Minnesota had an existing database that used the National Inventory of Wetlands (NWI) limits to delineate the lake/ponds. This example is used as a basis for “best practices” in associated documents.
The Capture Conditionshowever, were based on map and scale considerations. The lake/pond template includes two rules. The first says to collect ALL lake/ponds in an arid area, as defined in the instructions. The second rule says to collect lake/ponds that are at least 100 ft (.05in x 2000) in the shortest dimension in non-arid areas. This was both a cartographic and a resource decision. The printed line weight for the shoreline could take up 32 feet by itself (.008in x.008in x 2000), leaving very little space for the symbol fill. Anything smaller would be illegible on the map. In arid areas, because water is significant, some small lake/ponds were actually exaggerated on the printed map.
Source Interpretation Guidelines
Data Stewards and updaters need to provide similar information found in the Source Interpretation Guidelines. If the information already in the templates is acceptable, nothing needs to be done. However, if data has been revised using significantly different rules, then that methodology should be described in the Metadata.
Methodology for Identifying Features
Excerpt from: United States Geological Survey, 1990, An Enhanced Digital Line Graph Design, U.S. Geological Survey Circular 1048, p. 9-13
TEMPLATE DEVELOPMENT AND USE
About Feature Templates:
This information was documented as part of a comprehensive effort in the mid 1990’s to develop automated rules for USGS data revision and map production. The information was originally based on the content of 1:24,000-scale USGS topographic maps. Much of the information contained in this document is still valuable for describing an NHD dataset. It has been reformatted and is being included with the documentation for the NHD. This document will not be updated, but will evolve into a “best practices” document for NHD users and data stewards. Comments in BLUE have been added to clarify current practice; otherwise the document is unchanged from the 4/96 version.
Template Development and Use describes how to use the feature templates. The sections explain what is included in each part of the template and any global rules that apply throughout the templates. If a rule applies to all features, it is a global rule. An example is a rule for how to measure a feature to determine if it meets capture conditions. The templates contain only exceptions to the global rule or additional information that is unique to the feature. If nothing appears in the template to change a global rule, then the rule applies.
Throughout the templates, if something being described meets the definition and capture conditions of a feature, the feature name appears in all CAPS. Generic descriptions of features appear in lower case type to indicate they may not meet the definition of the feature and should not be considered as that feature.
The data extraction specifications in the feature templates contain all the information required to collect feature data. These specifications tell what is collected as a certain feature and when and how the feature is collected. The extraction specifications include:
Attribute and Attribute Values
Source Interpretation Guidelines
The templates are as concise and positive as possible, and each template is meant to stand alone. Therefore, definitions, attributes, and attribute values describe what the feature looks like, not what it DOES NOT look like. Similarly, capture conditions explain when to capture a feature, not when NOT TO capture the feature. If this approach had not been used, each template could be too cumbersome, and too confusing, to be useful.
The capture conditions reflect the NMD policy of cartographic collection, which restricts data content and position based on graphic limitations of scale and legibility. Only content that can be displayed on the printed map is collected and all offsets in position necessary to accommodate symbolization are performed at the time of data collection. This policy, and the feature templates, ill change as user requirements for geographic content and position become known and as technology and resources are developed to support the implementation of product generation rules for content generalization and symbol conflict resolution.
This requirement for “cartographic” collection will change as data stewards begin to maintain the NHD for a variety of uses, but the necessity to describe what was collected will not.
Feature definitions are used to decide how to classify a feature. Attributes, delineation, and capture conditions limit which occurrences of a feature, from a class of features, the NMD collects. The main goal in classifying features is to define the features so that the distinctions between them are clear. The features and their definitions were developed by studying a variety of sources including; NMD documentation; the Defense Mapping Agency's Feature and Attribute Coding System; Geographic Names Information System feature classes; the Spatial Data Transfer Standard feature list; publications from other Federal agencies, including National Ocean Service, Bureau of Land Management, Forest Service, and the Fish and Wildlife Service; and the Canadian National Topographic Data Base feature list. Attempts were made to coordinate feature definitions with other organizations; however, feature selection is somewhat different from one agency to another and even between units within each agency.
The feature definitions provide the distinguishing characteristics needed to differentiate between features. Although the difference between STREAM/RIVER and LAKE/POND is obvious, the distinction between STREAM/RIVER and CANAL/DITCH may not be so obvious. In this example, a feature that could be either a STREAM/RIVER or a CANAL/DITCH can be classified by comparing the two definitions. Although both STREAM/RIVER and CANAL/DITCH are linear water bodies, the definition for CANAL/DITCH specifies that it is artificial and that it is used to transport water, to drain or irrigate land, to connect two or more water bodies, or to serve as a waterway for watercraft. Therefore, CANAL/DITCH is distinguished from STREAM/RIVER by (1) the fact that it is artificial, and (2) the fact that it has specific uses. If the feature in question does not meet these two criteria, it is not a CANAL/DITCH.
Although the feature definitions include those characteristics of a feature that the NMD uses to distinguish among features, the templates do not necessarily specify how to make the distinction. How one goes about deciding if something is artificial or natural or if it is used for some special purpose or not, is beyond the scope of the templates. Annotation guides can be developed to support the content of the templates. These guides could contain graphic examples that illustrate map and real world identification and delineation of features.
There are some cases where the distinction between features is not clear, usually because past practices do not lend themselves to the classification method used to develop the domain of features. There are also cases where the definition is clear, but, again, because of past practices in the NMD, there might be some confusion. For example, the definition of a LAKE/POND states that it is a "body of standing water," so a dry lake doesn't fit the definition. However, the NMD symbol books describe dry lakes under lakes and ponds and in the DLG-3 format they are collected as lakes with a descriptive attribute of dry. In this case, a rule is developed in Source Interpretation Guidelines to reinforce the definition. The rule in this example is: “Do not capture dry lakes as LAKE/POND. See PLAYA.”
Attributes and Attribute Values
Attributes describe characteristics of features. Many of these characteristics fall into one of three groups: (1) "Type" describes the function or purpose of a feature; (2) "Category" describes the form or nature of a feature; and (3) "Status" describes the state or existence of the feature or characteristic.
Definitions for attributes and attribute values are generic. The definition for the attribute "Elevation" is "The vertical distance from a given datum." This applies whether elevation is applied to a LAKE/POND or a STREAM/RIVER or a CONTOUR.
In some cases, more than one value for a given attribute can be selected. The ability to provide multiple values for an attribute makes it unnecessary to capture multiple features. For example, if a mine produces multiple products, only one instance of the feature mine is captured and the applicable products are assigned as values to the attribute Product. Currently, the templates do not identify those attributes that can be multi-valued, although the information is stored in the standards data base. This requirement for multi-valued attributes was dropped because of limitations in GIS software.
For most features, there is a discrete list of appropriate attribute values. However, for a few features, such as RESERVATION, the number of potential descriptors is quite large and it is not possible to create an exhaustive list of values. Selecting an alphanumeric value for the attribute "Text" provides the necessary flexibility to describe a RESERVATION. No features in the NHD have this requirement for a “text” value.
Delineation specifications describe what the limits of a feature are and what to include in the feature that meets capture conditions. The delineation generally describes real world entities.
The representation rules are described in two tables. The first table lists the relationships in which a feature may participate and the second table lists the feature object types used to represent the feature.
The relationship table presents the relationship name, the cardinality, and the related feature object. The cardinality expresses the minimum and maximum number of times one instance of a feature can be involved in the relationship. However, the current design of the standards database is flawed and the cardinality cannot be entered properly, so the cardinality is not populated.
The representation conditions table presents the feature objects used to represent a feature and the criteria to determine which feature object is used. The values displayed in the columns for "AREA", "SHORTEST" and "LONGEST" are sizes based on an areal measurement, the shortest axis, or the longest axis of the feature. A feature is represented by a specific feature object when the size criteria in the appropriate column are met. If a feature can only be represented by one feature object, then the only value shown will be ">0" in one of the columns. If no values appear in any column, then special conditions must be present to indicate the appropriate feature representation. Special conditions may also exist in conjunction with values in table.
These tables were removed in the reformatted version of the NHD standards. The actual geodatabase implementation of the features class – for example NHDPoint and NHDLine – is described in the standards and listed after the feature definition. The conditions under which a feature is represented as a point versus a line, for example, were moved to the Delineation section.
The feature definitions describe what to capture, and the capture conditions describe when to capture it. Capture conditions are generally independent of source. The capture conditions currently reflect the content of a standard update product. Because primary mapping has been completed for the entire United States, most National Mapping Program activity is focused on revision. Information on data capture that pertains to specific sources or revision methods is found in the Source Interpretation Guidelines section.
The templates must contain the criteria necessary to ensure that NMD products are accurate and consistent in style and content. Therefore, the capture conditions present the requirements for the content of NMD products, not just the step-by-step decisions a user needs to make in deciding whether to capture a particular feature.
An “If...Then” format is used for the capture conditions. The basic format is as follows:
If FEATURE is CONDITION,
When there are multiple capture conditions, each statement stands alone. If the feature meets one of the conditions, it is captured.
Capture conditions are given in inches at map scale (ground distance in feet can be obtained by multiplying the numbers by 2,000). In general, features are measured along the longest axis (length) and/or the shortest axis (width). Square features are measured along either axis, round features are measured by the diameter, and irregular features are measured against the axes of the best fitting rectangle (non-oriented). Linear features are measured as the accumulative measurement along the centerline of the feature for length and the predominant distance across the feature (measured perpendicular to the centerline) for width. Any specific or unique requirements for measurement are addressed separately for each feature.
For areal measurements, the capture conditions are expressed as X square inches. Square inches indicate that the value is an areal value. Thus, 4 square inches indicates an area equivalent in extent to a square which measures 2" by 2", or to a rectangle which measures 1" by 4.
Decimal values are handled in the same way; an area described as 0.01 square inches indicates an area equivalent in size to a square which measures 0.1" by 0.1".
Do not confuse this terminology with the usage "2 foot square”, where neither the value nor the unit is areal.
The areal value terminology is used in the templates because it allows an area to be defined independently of lengths and widths. If there are minimum length or width requirements, then these values are included in the capture conditions in addition to the area value.
Attribute Information describes how to value the attributes once the feature is captured. Any required conditions and/or attribute value combinations are given. All attributes must be valued. There are three global attribute values that apply to many attributes. These are "Unspecified", "Not Applicable", and "Unknown".
· Unspecified is used when the value is not known, but is not necessary. For example, a spring shown on the map with no additional label would have the value Unspecified for the attribute Water Characteristics.
· Not Applicable is used when a particular occurrence of a feature cannot have a particular attribute value. For example, if the water level of a STREAM/RIVER is not controlled for navigation, the value for Elevation = Not Applicable, because the attribute does not apply and therefore cannot be valued.
· Unknown is used when a required value is not known. For example, if the class of a road cannot be determined during collection or revision, the value would be unknown. Other sources will be required to determine the appropriate value.
Source Interpretation Guidelines
Source Interpretation Guidelines provide additional information for interpreting the capture conditions when capturing data using specific source materials or methods. They also contain any modifications to the capture conditions specific to the source or capture methods.
· All sources
This information helps interpret the capture conditions regardless of the source or method used in data capture. Included are such things as when to capture a coincident feature based on the capture conditions, when to capture more than one instance of the feature, and when to capture something as a different feature instead of the feature in the template. For those features that can occur in more than one theme, this section provides the guidelines for which theme should contain a specific feature instance.
This information helps interpret the capture conditions when the source is a map. Included are guidelines for interpreting the symbology for proper classification, delineation, and capture. When deciding to capture a feature from a graphic source, the capture conditions still apply. Features that do not meet the capture conditions are not captured. Generally, this reflects changing requirements. For example, a number of offshore features shown on NMD maps prior to 1961 are no longer required and should not be captured, even though they appear on the graphic.
Some capture conditions cannot be evaluated just by looking at the map. For example, when a feature is represented with a point symbol and the capture conditions state a size requirement, it is not possible to evaluate the true size of the feature from the graphic. If compliance with the capture conditions cannot be determined, then the feature is collected. Further evaluation will be done at the time of revision.
In some cases, instructions are given to collect features for which the symbology has been suppressed on the map. For example, instructions are given on how to capture PLSS information when PLSS lines are dropped from the map because they are coincident with a boundary or a road.
This information helps interpret the capture conditions during revision. The term "revision" applies to the process by which data are updated to reflect changes that have occurred since the date of the existing DLG or, for simultaneous collection and revision, the Digital Raster Graphic (DRG).
Guidelines in this section are divided by the category of revision. Headings in this section are:
Revision - General
Revision - Standard
Revision – Limited
If no guidelines appear in any of these revision categories, then the guidelines in the remainder of the feature template apply.
Revision - General
Revision - Standard
Revisions - Limited - The goal for a limited update is that feature content will be current, but will include only: (1) those feature types that are photoidentifiable on a monoscopic source, supplemented with limited ancillary sources, and (2) those feature types from existing DLG's or DRG's that are not photoidentifiable but are not particularly prone to change. Some feature types are not revised at all while other feature types are revised with limited attribution. Existing data may or may not be revised. Ancillary sources may be required to revise some data. All of this information is provided in the limited update section.
If no limited update guidelines appear, the feature and all its attributes are revised using guidelines in the remainder of the template.
If a feature is not revised during a limited update, guidelines state this and give information on how to handle existing DLG data. The guideline "Do not revise. Delete existing features." means the feature is not revised and all feature instances are deleted from the DLG. The guideline “Do not Revise. Retain existing features.” means the feature is not revised and features that already exist in the DLG are retained except when that feature is replaced by another instance of a feature type that is revised in a limited update.
If a feature is revised, but only under certain conditions, those conditions will be stated, as well as how to handle existing DLG data. For example, in the feature PIPELINE, the guideline states “Revise aboveground pipelines only. Retain existing features.”
There are some cases where only existing DLG data are revised and only under certain conditions. The guideline "Do not add new features. Revise existing features (conditions)." applies to these cases.
Special instructions may also apply to attribution. If attribution cannot be determined in limited update, that will be noted in this section. For example, the
guideline for BUILDING states that Building Type = Unspecified for limited update. In addition it will be noted whether the value applies to existing data or
not. In the case of BUILDING, the guideline states “Existing buildings will be given Building Type = Unspecified.”
If data is being collected from a DRG as part of a simultaneous collection/revision project, the limited update instructions still apply. When revising data from a DRG, the term "retain" is interpreted to mean "collect" and the term "delete" is interpreted to mean "do not collect". So, the guideline "Do not revise. Retain existing features" means "Do not revise. Collect existing features". The guideline "Do not revise. Delete existing features" means “Do not revise. Do not collect existing features.”