by contractors or airport staff (collectively referred to as “Geospatial Data Producer”) when developing geospatial data from field observations. The Geospatial Data Curator will coordinate all geospatial data collection conducted by the airport. Whether data collection is to be performed outside or inside of buildings will dictate what processes and equipment the Geospatial Data Producer should use.

The Geospatial Data Producer will choose a data collection device that corresponds to the type of data to be collected, the accuracies required, and accessibility of the features being recorded. For exterior work, redlining on hardcopy or electronic maps, survey-grade real-time kinematic (RTK) GPS units may be selected. For interior data collection, redlining on hardcopy or electronic maps or electronic measuring devices (such as a distometer laser measuring device or lidar scanner) may be selected. Regardless of where the data is to be collected or the type of data collection method/device the Geospatial Data Producer plans to use, the next step is to download or print existing information.

For exterior data collection, the Geospatial Data Producer should make redline annotations to hardcopy or electronic maps used for exterior data collection to record observations in the field. Hardcopy redlines used for exterior data collection should be transposed into an electronic format by the Geospatial Data Producer once they have returned from the field. Points, along with any associated attributes and metadata, should be recorded in the field. Points captured using GPS may need to be post-processed to obtain the desired accuracies. For interior data collection, the Geospatial Data Producer should make redline annotations to hardcopy or electronic maps used for interior data collection to record observations in the field.

Hardcopy redlines used for interior data collection should be transposed into an electronic format by the Geospatial Data Producer once they have returned from the field. Interior measurements may be taken using measuring tape or electronic devices. The next step, regardless of how field data is collected, is to convert the information into a format that can be used within the data warehouse to support geometry and attribute data maintenance. Data collected in the field should be checked by reviewing error statistics from GPS equipment, randomly tested against another source of equal or higher accuracy, compared with high-resolution orthophotos, and through other means.

Electronic Data Creation

The following steps describe how new data that is collected should be assessed and prioritized for incorporation into the data warehouse. These assessment procedures evaluate the quality of the new data in comparison with the current data. Quality is assessed based on the accuracy, correctness, comprehensiveness, and other factors. Where the new data is determined to be of higher quality than data currently in the data warehouse, this assessment will help prioritize the efforts required to clean up, convert, and properly load it into the data warehouse.

The first step in data assessment is to review any metadata that may have been provided with the source data. If no formal metadata (e.g., metadata saved within a data warehouse or as an XML component) is provided, information including transmittal letters, labels, “readme” files on the media and other supporting documentation should be reviewed. If this information is still insufficient, then the person who provided the data should be contacted for further details.

The metadata should be checked for temporal accuracy (i.e., how current it is). If the data is determined to be newer than data in the data warehouse, then it may be of use. The data should be checked for positional accuracy (i.e., how close it is to the feature’s true position). If the data is determined to be more accurate than data in the data warehouse, then it may be of use. The data should be examined to determine the physical extent of the data to ensure that it is relevant to the airport.

planimetrics carried out by the airport is an important source of wholesale feature data updates. Geospatial Data Producers will also record a variety of data in the field. Finally, other divisions may report discrepancies between existing data and current conditions in the field. Sometimes, these discrepancies may prompt further fieldwork to investigate and resolve any discrepancies between the geospatial data and the actual field conditions.

Regardless of the source of the data to be used for data maintenance, the Geospatial Data Producer will first determine if this data complies with the airport’s standards. In most cases, data will be coming from the airport staff or consultants and must comply with the standards. In cases where data is coming from third-party vendors, municipalities, or other sources, compliance with airport standards may not be required.

The next step is for the Geospatial Data Producer to determine what type of data has been received. CAD data may need to be converted into a format compatible for loading into a GIS, or vice versa. Survey data also may need to be converted into a format that can be edited and prepared for loading. Many formats such as Comma Separated Value (CSV) coordinates and Coordinate Geometry (COGO) files can be read directly by GIS and CAD software. Other formats, such as LandXML, may require conversion.

With all incoming data in a format that can be edited in the airport, the Geospatial Data Producer should determine if the new data is expressed in ground- or grid-based coordinates. Although the airport standards may require data in grid-based coordinates, some data may be submitted by sources that do not have to adhere to these standards and may be in ground-based coordinates or a local assumed coordinate system. In many cases, metadata will be provided that will indicate which coordinate system was used. In other cases, features in the incoming data may have to be compared with the same features in a known coordinate system.

If the incoming data has been determined to be in ground-based coordinates, the ground/grid scale factor must be applied using the appropriate coordinate transformation functions. If the incoming data has been determined to be in a local assumed coordinate system, the ground/grid scale factor must be applied using the appropriate coordinate transformation functions and the data will need to be translated using common features.

The Geospatial Data Producer then loads the data into a location where a working copy should be created for evaluation, leaving the original undisturbed. Software data management tools may also facilitate creation of an “edit” version of the data. The Geospatial Data Producer can then begin the process of reviewing the incoming data to determine what changes need to be made to the data warehouse. In many cases, the new data will contain new features that do not exist in the data warehouse and must be added.

The Geospatial Data Producer may find that new features, created by others, must be cleaned before being added to the data warehouse. Following are the typical types of issues that will need to be cleaned from new data sets:

• Polygons may not be closed. In these cases, polygons should be closed by snapping two existing vertices together or by adding a new line segment that closes the gap. If the polygon is not closed, it will not be represented as a polygon and will need to be closed for loading into the data warehouse.
• Polylines may need to be manipulated so that each polyline or line segment connects to adjacent nodes (typically shown as point features).
• Polygons that fall below a logical size limit may have been created as an anomaly in the data creation process and may need to be removed.
• Vertices of polygons and lines should be sufficient to achieve the desired accuracy. Additional vertices that do not provide clarity to the data could be removed using various software tools.

specifications for preparing compliant CAD data are described in the CAD standards. Existing CAD data sources or CAD data that is not prepared by the airport staff or consultants will not conform to the airport CAD standards. In these cases, CAD data may need to be manipulated by the Geospatial Data Producer before it can be incorporated into the airport’s data warehouse.

CAD Preparation and Conversion Steps

Where applicable, check for compliance with the airport’s data standards. Each CAD, DWG, or DGN file should be checked to ensure that it can be opened in the correct version of Autodesk AutoCAD or Bentley’s MicroStation, respectively. The model space of each CAD drawing should be checked to ensure that it is in the proper geographic coordinate space. This can be accomplished by checking the drawing coordinates of at least two points shown in the CAD drawing against the true known coordinates for those points. Control stations, for which high accuracy coordinates are known, are recommended for this purpose.

The Geospatial Data Producer should randomly check a representative sample of layers which will be used for feature creation, modification, or archived to ensure that the proper geometry and type of features are on each layer.

Depending on the type of data and the number of similar CAD files to import, the Geospatial Data Producer may decide to use software tools to convert the CAD data into a data warehouse format. FME or other data conversion software could be considered in the following cases:

• CAD files to be converted contain a large number of layers for which there are many CAD layers that are to be combined into a single GIS layer with an attribute (typically a type code) that distinguishes between them. FME’s workbench can be used to read the layers in the CAD file, and they can be graphically matched with the corresponding layer(s) of the GIS. The separation of features on different CAD layers can be used to set an attribute in the GIS file.
• Multiple CAD files are submitted with the same layers but for different geographic areas. An example of this is when sections of a building are split into separate files with match lines to align the data. FME Workbench can be used to create a translation file that the Geospatial Data Producer can save and reuse for each CAD file of the same structure.
• CAD files are submitted as part of a long-term project where future submittals are likely to require the same layer manipulation. FME Workbench can be used to create a translation file that the Geospatial Data Producer can save and reuse for future CAD files of the same structure.

The technician should then bring the data from the CAD file or from the file(s) created by FME into the data warehouse for manipulation and editing. The geometry maintenance procedures should now be carried out with the CAD data.

Attribute Data Maintenance

This procedure describes methods for the entry, cleanup, conversion and loading of alphanumeric data that, in some cases, is maintained separately from the corresponding geometry in a data set. This procedure covers text that is manually entered and text that is loaded from electronic files—it does not include procedures for transferring data from other systems.

Attribute Data Maintenance Steps

The data creation process will produce attributes that are usually entered by Geospatial Data Producers or external data providers and then collected by the Geospatial Data Curator. Attributes will also be produced through field data collection activities. This field data collection will be coordinated by the Geospatial Data Curator. The first step in the attribute maintenance

process is for the Geospatial Data Producer to determine the format of the attribute data to be entered into the GIS. Viable attribute formats can include hardcopy data, spreadsheets (e.g., Microsoft Excel), personal databases (e.g., Microsoft Access), and comma delimited text files.

Attribute data provided in a hardcopy format will need to be manually entered into the attribute table for the appropriate feature class within GIS. In some cases where a large number of attributes are to be entered, the Geospatial Data Producer may decide to enter the data into another format such as a Microsoft Excel spreadsheet, which can allow long lists of data to be more easily copied and checked, then joined to the appropriate feature table.

Attribute data provided in tabular electronic format such as a Microsoft Excel spreadsheet, Microsoft Access database, or CSV ASCII file can often be linked with features in a spatial data set. For this to work, the feature class and the external tabular data to be joined need to share a common unique identifier. This can be achieved by exporting attribute data including such an identifier from the feature class first and then using this exported file for subsequent data entry. When the external data is joined to the feature class, the joined attributes are not saved into the feature class directly, but they can be copied into the desired attribute fields.

In some cases, attributes may be provided in GIS or CAD files in a manner that can be converted to the appropriate attributes in the data warehouse. Tools such as FME Workbench and others can help in such conversion. For example, GIS files may contain fields that need to be concatenated or edited before being compliant with the GIS Spatial Data Standards Specification. With CAD data, attributes can be attached to blocks or as object data. In some cases, tools such as FME can be used to convert these attributes into a format that can be read by the GIS.

After attributes are entered, metadata should be adjusted as necessary to reflect the changed entries. Metadata changes may include the source of the revisions, the date the edits were made and any notes regarding cleanup or conversion steps that may have been used to alter the data.

QA should be completed on all modified data sets once the modifications are completed by the Geospatial Data Producer. After all modifications have passed QA, the Geospatial Data Producer should post the new data to the production data warehouse.

GIS Data QC, QA, and Validation Procedures

This section describes QC, QA, and validation procedures that are to be applied to data that is to be loaded into the GIS. These procedures include those that the airport personnel responsible for data maintenance should use as data is being maintained. Consultants and contractors that develop data for the GIS should also be encouraged to use these procedures to ensure data will pass the airport QC checks. The procedures in this section also include steps that the staff should use to check contractor data submittals against the data standards and specifications before it is loaded into the data warehouse. Finally, the process for carrying out annual audits to check data in the data warehouse is described.

QC Procedures

The procedures that data maintainers should use to establish quality as they develop and maintain data are described in this section. These procedures are mainly defined for the airport staff responsible for data maintenance, but consultants and contractors should be encouraged to use them as well. Consultants and contractors cannot be mandated to use these procedures as they may have designed procedures specific to their business process.

As geometric data is created, it should be checked for spatial accuracy. This is especially the case if data is being created based on source data that is converted from another coordinate

system such as ground coordinates. Data can be checked for accuracy by comparing it with a source of similar data that is known to be of higher quality. Two sources that can provide such information, depending on the expected quality of the data being checked, are the planimetric data provided as a result of the orthophoto development process or fieldwork to collect high-accuracy coordinates.

As features are created, edited, or archived, the Geospatial Data Producer should check to ensure that all pertinent features are properly handled. There are several ways this can be done depending on the data. A query to indicate the number of features edited can be executed and compared with the number of features in the source data set. A visual review of the physical extent of the area being edited can be conducted. Edited features can be overlaid on top of aerial photography or features from the data warehouse to ensure all expected features have been edited. In some cases, field verification of the data may be warranted.

All features created or edited should be checked for conformance to the appropriate geometric rules. Although these will vary between feature classes, the following general rules apply to all based on their geometry type:

• Polygons may not be closed. In these cases, polygons should be closed by snapping two existing vertices together or by adding a new line segment that closes the gap.
• Polylines may need to be dissolved so that each polyline or line segment connects to adjacent nodes (typically shown as point features).
• Polygons that fall below a logical size limit may have been created as an anomaly in the data creation process and may need to be removed.
• Line segments and polylines may need to be merged so that each polyline or line segment connects to adjacent nodes (typically shown as point features).
• Vertices of polygons and lines should be sufficient to achieve the desired accuracy. Additional vertices that do not provide clarity to the data should be removed.

All features created or edited should be checked for conformance to the appropriate topology rules. Although some rules are specific to individual feature classes, the following rules apply to all:

• Polygons may overlap each other in cases where overlaps are not allowed (e.g., lease lines). In these cases, polygon edges should be adjusted to remove the overlap.
• Polygons may have slivers of unaccounted space between them. In these cases, polygon edges should be adjusted to close the gap.
• Line overshoots may need to be trimmed so that line segments end at a node, endpoint, or point of intersection as appropriate.
• Line undershoots may need to be extended so that line segments end at a node, endpoint, or point of intersection as appropriate.
• Vertices of features should be snapped to corresponding vertices of adjacent features.

As attributes are entered, they should be checked for correctness against the source data and conformity with the format and rules relevant to each attribute. For example, all numbers should be checked to include only numeric values, positive values should be positive, dates should be valid dates, etc. Software validation routines will check for many of these issues, but manual QC should also be performed. One method of doing this is to sort attributes in a table and examine the results for outliers. The Esri “frequency” tool is also useful for examining similarly attributed features for possible attribution errors.

Metadata should also be checked in much the same manner as attributes are checked. The Geospatial Data Producer should proofread all narrative and explanatory text in the metadata to ensure that it is logical and does not contain grammatical or typographical errors.

control and quality assurance procedures, and then resubmit the data for validation (now at a lower threshold of failures).

Data Audit Procedures

After new data sets are introduced, the airport configuration is updated, or a significant period passes; periodic audits should be performed to check data in the data warehouse against actual conditions in the field, including looking for geometry and attributes that may have changed. A periodic orthophotography update is also an audit in that it can be used to compare current conditions with existing data that will then be brought up to date.

Operations and maintenance personnel are constantly in the field making observations and should report any discrepancies that they observe between current conditions in the field and the map data that they are using. Although less formal, these ongoing observations are an important way of keeping up with the dynamic nature of airport infrastructure changes. Regardless of the data being checked or the frequency of the audit, these data audits are a form of field data gathering and should be carried out as described earlier. In summary

• Current data should be downloaded or printed from the data warehouse;
• Field observations should be made comparing current conditions with the data;
• Redlines should take note of any discrepancies (in some cases, measurements or coordinates will be taken); and
• The data recorded in the field should be converted into the data warehouse upon return.
• Finally, the Geospatial Data Producer responsible for the data should modify geometry and/or attributes, as appropriate.

References