Differences in NHN Data

This section addresses known particularities, deviations or differences in some NHN data as opposed to the National Hydro Network standards and specifications. Since NHN data are produced using the best available source data - provincial data, for example - they de facto inherit from them. NHN data may, as a result, reflect source data limitations or constraints.

Particularities in NHN data produced by Natural Resources Canada (NHN-CL4)

NHN data products produced by Natural Resources Canada (NHN-NRCan) following its initial NHN data production process (2007), namely NHN data of the Completeness Level 4 (NHN-CL4), do have the following particularities (in their Distribution Profile) compared to the initial NHN data model. For more information on the initial NHN data model, please refer to the National Hydro Network, Canada, Level 1, Data Model, Edition 1.0, 2004-08 document, which can be found under the NHN data Documentation section under NHN Standard (conceptual description):

  1. [Water Definition attribute] Attribute values "DITCH" (3), "TIDAL RIVER" (7) and "SIDE CHANNEL" (10) are not present in NHN data initially produced by Natural Resources Canada (NRCan).

Particularities in NHN-CL1 data produced by Natural Resources Canada

NHN data products produced by Natural Resources Canada (NHN-NRCan) following its NHN-CL1 data production process, namely NHN data of the Completeness Level 1 (NHN-CL1), fully comply to NHN-CL1 minimal requirements. However, those NHN-CL1 data have the following particularities (in their Distribution Profile) compared to the initial NHN data model. For more information on the initial NHN data model, please refer to the National Hydro Network, Canada, Level 1, Data Model, Edition 1.0, 2004-08 document, which can be found under the NHN data Documentation section under NHN Standard (conceptual description):

  1. [Littoral, Bank] NHN-CL1 data (Distribution Profile) do not contain the "Littoral" feature initially planned for in the NHN Hydro Network package. Actually, this feature has been replaced by the "Bank" feature for which the "WATER DEFINITION" attribute has the value "UNKNOWN" (-1).

  2. [Named Feature] The NHN Toponymy package (Distribution Profile) contains a single feature, the "Named Feature ", which can have a point-type, line-type or area-type geometric representation. According to the National Hydro Network Feature Catalogue, Distribution Profile, Edition 1.0, 2007-06-01 document, this feature delimits, in whole or in part, a toponymic phenomenon whose geometry does not fully match the geometry of a feature from the NHN Hydrographic package. Now, in NHN-CL1 data, the feature "Named Feature" has only one single point-type geometric representation. Moreover, the Named Features found in NHN-CL1 data carry toponyms or geographical names that fall in three different categories, that is to say : 1- Toponyms that point out to toponymic phenomena whose geometry does not fully match the geometry of a feature from the NHN Hydrographic package; 2- Toponyms that point out to toponymic phenomena whose geometry does fully match the geometry of a feature from the NHN Hydrographic package, but whose automatic correlation with the corresponding NHN geometric feature has not been completed; 3- Toponyms that point out to coastal toponymic phenomena located in the sea.

  3. [Waterbody] NHN data (Distribution Profile) do contain the "Waterbody" feature initially planned for in the NHN Hydrographic package. However, in the case of NHN-CL1 data, it is frequent that this feature bears the value "UNKNOWN" (-1) for its "WATER DEFINITION" attribute. A direct consequence from this happens in coastal areas where some waterbodies do extend into the ocean, whereas they should stop at the littoral according to the initial NHN data model.

  4. [Hydro Events package] NHN-CL1 data contain point-type and line-type "Manmade Event" and "Obstacle Event " features. According to the initial NHN data model, NHN events are created from the geometric intersection or geometric projection of a "Manmade Hydrographic Entity" or "Hydrographic Obstacle Entity" with a network linear element (Network Linear Flow or Bank). However, in NHN-CL1 data, all events found result from a geometric intersection only.

  5. [Entity Type attribute] The feature "Named Feature" from the Toponymy package has an "Entity Type" attribute, which expresses the nature of the named topographic phenomenon. The attribute value is always set to "Inland Water" (8) in the NHN. However, in NHN-CL1 data, the "Named Feature" is used in ways that do not always comply with the feature's definition (refer to preceeding explanation about the "Named Feature" characteristics for NHN-CL1 data). In particular, this feature is used to carry toponyms or geographical names that relate to coastal toponymic phenomena located in the sea. In those very specific cases, the "Entity Type" attribute value should be "Marine Waterbody" (1) or "Undersea Feature" (7), which both represent entities or named phenomena located in the sea. Since the only value found in NHN-CL1 data for the "Entity Type" attribute is always "Inland Water" (8), it entails a difference or deviation as opposed to NHN standards and specifications. For more information, refer to the National Hydro Network Feature Catalogue, Distribution Profile, Edition 1.0, 2007-06-01 document.

  6. [Flow Direction attribute] The "Flow Direction" attribute used to tell the water flow direction compared to the "Network Linear Flow" digitizing direction is dealt with in a very specific manner in NHN-CL1 data. Only the "Network Linear Flow" features having the value "PRIMARY" (1) for the "Level Priority" attribute and the value "FALSE" (0) for the "Isolated" attribute are oriented (digitizing direction) downstream and, therefore bear the value "Same Direction" (1) for the "Flow Direction" attribute. "Network Linear Flow" features having the value "TRUE" (1) for the "Isolated" attribute or located within "Waterbody" features whose "Isolated" attribute is set to "TRUE" (1) bear the value "N/A" (3) for the "Flow Direction" attribute, whereas all others bear the value "UNKNOWN" (-1).

  7. [Level Priority attribute] The "Level Priority" attribute is used to classify the route or path followed by the Network Linear Flow feature within the NHN hydrographic network. The domain of values for this attribute includes the following values : "UNKNOWN" (-1) ; "PRIMARY" (1) ; "SECONDARY" (2). The "PRIMARY" value identifies the main route whereas the "SECONDARY" value identifies an alternate or secondary route followed by the Network Linear Flow feature. However, in the NHN-CL1 automatic production process, the "PRIMARY" value is allocated based on the shortest path principle instead of the size and importance of the watercourse. Thus, in NHN-CL1 data, it is possible to have a narrow and shallow section of a watercourse be identified as "PRIMARY ", whereas a wider and deeper but longer alternate section be identified as "SECONDARY ".

Particularities in NHN-CL2 data produced by Natural Resources Canada

NHN data products produced by Natural Resources Canada (NHN-NRCan) following its NHN-CL2 data production process, namely NHN data of the Completeness Level 2 (NHN-CL2), fully comply to NHN-CL2 minimal requirements. However, those NHN-CL2 data have the following particularities (in their Distribution Profile) compared to the initial NHN data model. For more information on the initial NHN data model, please refer to the National Hydro Network, Canada, Level 1, Data Model, Edition 1.0, 2004-08 document, which can be found under the NHN data Documentation section under NHN Standard (conceptual description):

  1. [Named Feature] The NHN Toponymy package (Distribution Profile) contains a single feature, the "Named Feature ", which can have a point-type, line-type or area-type geometric representation. According to the National Hydro Network Feature Catalogue, Distribution Profile, Edition 1.0, 2007-06-01 document, this feature delimits, in whole or in part, a toponymic phenomenon whose geometry does not fully match the geometry of a feature from the NHN Hydrographic package. Now, in NHN-CL2 data, the feature "Named Feature" has only one single point-type geometric representation. Moreover, the Named Features found in NHN-CL2 data carry toponyms or geographical names that fall in three different categories, that is to say : 1- Toponyms that point out to toponymic phenomena whose geometry does not fully match the geometry of a feature from the NHN Hydrographic package; 2- Toponyms that point out to toponymic phenomena whose geometry does fully match the geometry of a feature from the NHN Hydrographic package, but whose automatic correlation with the corresponding NHN geometric feature has not been completed; 3- Toponyms that point out to coastal toponymic phenomena located in the sea.

  2. [Hydro Events package] NHN-CL2 data contain point-type and line-type "Manmade Event" and "Obstacle Event " features. According to the initial NHN data model, NHN events are created from the geometric intersection or geometric projection of a "Manmade Hydrographic Entity" or "Hydrographic Obstacle Entity" with a network linear element (Network Linear Flow or Bank). However, in NHN-CL2 data, all events found result from a geometric intersection only.

  3. [Flow Direction attribute] The "Flow Direction" attribute used to tell the water flow direction compared to the "Network Linear Flow" digitizing direction is dealt with in a very specific manner in NHN-CL2 data. Only the "Network Linear Flow" features having the value "PRIMARY" (1) for the "Level Priority" attribute and the value "FALSE" (0) for the "Isolated" attribute are oriented (digitizing direction) downstream and, therefore bear the value "Same Direction" (1) for the "Flow Direction" attribute. "Network Linear Flow" features having the value "TRUE" (1) for the "Isolated" attribute or located within "Waterbody" features whose "Isolated" attribute is set to "TRUE" (1) bear the value "N/A" (3) for the "Flow Direction" attribute, whereas all others bear the value "UNKNOWN" (-1).

  4. [Level Priority attribute] The "Level Priority" attribute is used to classify the route or path followed by the Network Linear Flow feature within the NHN hydrographic network. The domain of values for this attribute includes the following values : "UNKNOWN" (-1) ; "PRIMARY" (1) ; "SECONDARY" (2). The "PRIMARY" value identifies the main route whereas the "SECONDARY" value identifies an alternate or secondary route followed by the Network Linear Flow feature. However, in the NHN-CL2 automatic production process, the "PRIMARY" value is allocated based on the shortest path principle instead of the size and importance of the watercourse. Thus, in NHN-CL2 data, it is possible to have a narrow and shallow section of a watercourse be identified as "PRIMARY ", whereas a wider and deeper but longer alternate section be identified as "SECONDARY ".

Deviations in NHN-CL4 Data from British Colombia

British Colombia NHN Data (NHN-BC) present the following deviations as opposed to the NHN standards and specifications:

  1. [Manmade Features] The BC-NHN does not contain Lock Gate, Boat Ramp, or Fish Ladder manmade features.

  2. [Obstacles] The BC-NHN does not contain Reef, Rocks, Exposed Shipwreck or Ford features obstacle features.

  3. [Network Linear Flow] The BC-CWB dataset contains culverts, storm sewers, and other inferred edges used for stream connectivity. These features are not observed and will not be present in the NHN Hydrographic package, however these edges are present in the NHN Network Linear Flow Class with a Network Flow Type of Constructed.

  4. [Obstacles and Manmade Features] The BC-NHN contains dam, falls, dock and wharf features which may exist as "Points" or "Lines" inside of a waterbody. They may also exist on isolated single line watercourse features.

  5. [Toponymy] In BC-CWB the toponymic information of named streams is only captured on the stream network. However in NHN this information is captured on both the network and hydrographic packages. Because of the constructed features in the CWB dataset there is a many-to-many relationship between the network linear flow and single line watercourse features. This many-to-many relationship results in single line watercourse features that are comprised of multiple network linear flow features of which only parts may be named. In BC-NHN, however, the entire single line watercourse feature is named which results in toponymic inconsistencies. For accurate toponymic information, the network linear flow named features should be used.

  6. The BC international or provincial/territorial boundary does not match the GeoBase geopolitical boundaries along the BC-Alaska and BC-Washington, BC-Idaho, BC-Montana or BC-Alberta borders. The differences vary on either side by up to 2 km with the largest differences occurring along the BC-Alaska and BC-Alberta Borders

  7. [Manmade Features] The mapping of BC DryDock, FerryDock, MarinaDock, Pier/Wharf to NHN manmade features is as follows:

    • BC DryDock = NHN Slip
    • BC FerryDock = NHN Wharf
    • BC Marina Dock = NHN Wharf
    • BC Pier/Wharf = NHN Wharf

  8. [Manmade Features: Wharf] The wharf manmade features were generated with the following additional spatial constraints: wharf CROSSES littoral OR relate (wharf, waterbody, FTT******).

  9. [Events] The BC-NHN contains only events for manmade and obstacle features that are projected less than or equal to 10 metres.

    BC manmade events for non-dike entities were generated by:

    1. Creating events for intersections between manmade entities and associated network linear features:

      • A linear event was created for each network linear feature from the two further intersection points along the feature.
      • A point event was created if only a single intersection existed for the network linear feature (note that multiple point events will occur at the same x,y location if the entity intersects the network linear feature at a node (junction)).

    2. Creating projected events for remaining entities.

      • For point entities, point events were created by using the nearest point along a network linear element within 10m.
      • For linear entities, the entities were intersected with all network linear features buffered by 10m. For each resulting intersecting feature, the start and end point of the geometry were projected back to the network linear feature and a linear event was created using the two point furthest away from each other on the network linear features. If the resulting linear event was less than 5m it was converted to point event (in an attempt to meet the requirement that entities that are perpendicular to network linear features create point events).
      • For polygonal entities, point events were created on the nearest network linear element within 10m of the polygon entity.
      • For all features if line events were created, point events were not created.
      • Dike events were generated using an intersection between 100m buffer around the dike and all associated network linear features within 100m. Next, cases where the entire network linear feature associated with the event was greater than 10m away from the dike entity were removed. This implies that a dike entity may have multiple events along multiple banks or littorals and dike events may be created along the network linear feature where part of the feature is greater than 10m from the dike entity; however, some other part of the feature must be within 10m. The reason for doing this was that there are many dikes that move in and out of the 10m range which would have resulted in many tiny linear events which is not representative of the data.

    BC obstacles events were generated by:

    1. Creating events for intersections between obstacle entities and associated network linear features:

      • For linear intersections 2-vertex linear events were created.
      • For point intersections point events were created.

    2. Projected events were created for all entities without intersecting events and whose network linear element lies within 10m of the obstacle entity.

      • For point, line, and polygon entities, point events were created by using the nearest point along a network linear element within 10m

  10. [Manmade features] The BC-NHN contains one Dam feature which is associated with an isolated Single line watercourse.

  11. [Obstacles] The BC-NHN contains 11 Falls features that are associated with an isolated Single line watercourse.

  12. [Toponymy] The BC-NHN has not attached river names to lake features as per the NHN model.

  13. Along the BC / USA borders, BC has used extra-jurisdictional data when computing isolated attributes. Thus, some features along the border may be flagged as not isolated because data exists in the extra-jurisdictional regions.

The TRIM data product (which was used as input to the BC-NHN) contains the following specifications:

  1. The TRIM dataset was captured from stereoscopic photo interpretation with a horizontal planimetric accuracy of +/- 10m, 90% of the time, and a vertical accuracy of +/- 10m, 90% of the time.
  2. Any hydrographic feature is captured as linear when the feature is less than 20m wide, and polygonal when greater.
  3. The minimum size of a lake is captured where the longest dimension is over 25 meters.
  4. The maximum stream width at which the feature is captured as a line is 20 meters or less.
  5. The maximum width at which the canal is captured as a line is 20 meters or less.
  6. The coastline definition is captured at the high tide mark (high water level).
  7. The maximum width at which a dock or wharf is captured as a line is 20 meters or less.