Check descriptions - sections

Distance to Next vs Node Name

This can help trap errors caused for instance by extra sections (e.g. interpolates) being added, without account being taken of the reduced distance between sections. It relies on a node-naming convention which explicitly includes a cumulative distance for nodes in a single reach.

The check attempts to reconcile node names against the distance to next section. It will work for river sections, interpolates, conduits and replicate units. The user can specify an acceptable error range (default 0.750 m) in the “Dis Err” (m) box. The check can be performed in either metres or km depending on which unit is used in the node name. The check will then list any nodes with a difference between the node name and the distance to next greater than “Dis Err”. Alongside the node name is given the chainage derived from the node name, the chainage derived from the following node name, the distance to next and the difference between node name chainage. These are all given in metres.

For example if a model has node XX3400 with distance to next of 53 m and the following node is named XX3357, where node name chainage is in metres, the check would return an error:

Note - This check is only valid if the node naming strategy fits the programs methodology. In general a node name consists of [XX][CHAINAGE][CHAR] where:
XX is the river identifier
CHAINAGE is the chainage
CHAR is an optional additional descriptor that can contain letter or numbers.
The program would convert the following:

Nodes with Zero Initial Levels

This is a simple but common reason for a model to fail to run. Nodes may have been added to a model and initial conditions have been omitted.

This check looks for any nodes, with initial condition of zero and returns a warning with a list of nodes which have failed the check. Spills with zero initial conditions are ignored, but other structures will return a warning.

Nodes With Below Invert Initial Conditions

This is a common reason for a model failing to run. The check will compare the initial conditions with the lowest level in a river cross section or the invert level of a culvert. It will return a warning with a list of nodes which have failed the check and the corresponding invert level and initial stage alongside. Only river sections and conduits are checked, therefore structures need checked by hand and interpolates / replicates should be checked to ensure initial conditions have been graded between river / conduit sections.

Panel Marker Existence

Panel markers are required in river cross sections where a composite channel is required or where the Manning’s roughness changes within a cross-section or when the flood plain is represented using extended sections. The check looks for at least one panel marker per section. No attempt is made to ensure that the panel markers are in the correct location. If no panel markers are found in a river cross section an error will be returned and the node name listed.

Left / Right Bank Markers

Bank markers do not affect the model results, however they are useful for determining when a river section comes out of bank. The program looks for the left and right bank markers in river cross sections and if either marker is missing then a warning is made and the node name listed. No attempt is made to ensure that the bank markers are in the correct location.

River Sections Cut Back (With Spills)

When spills are attached to river sections, it is important that the cross section has been cut back to avoid double counting area. This checks river sections with spills attached to ensure the highest points of the cross section occur at the left and right extremes. When only one spill is attached the program will attempt to match the spill against the relevant channel side. If the left and right end points are not the highest points it reports the section as an error and which side it interprets as having a problem.

River Sections Cut Back (All Sections)

This check ensures that all river cross-sections in the model have the extreme end points higher than the rest of the cross-section. If the left and right end points are not the highest points it reports the section as an error and which side it interprets as having a problem. This is important as it could indicate the cross-section has not been extended realistically or sections have not been cut back enough.

River Section LiDAR / SAR extension disjoints

There is often a discrepancy between cross-section survey data and LiDAR / SAR data used to extend river cross-sections. It is important to reconcile the LiDAR / SAR, to account for its inaccuracy, with the cross-section survey to ensure accurate modelling of the floodplain. This check looks at every cross-section to find a point with a large jump in elevation (0.1 - 1.0 m) which occurs at a single point. It ignores the jump if the sections either side are flat or have a difference in elevation greater than 0.1 m as these are likely to be realistic situations, e.g. walls or steep slopes. If the program feels there is an inconsistency an error is reported and the node name is listed.

Glass Walling against Max Stage csv File

"Glass walling" is a potential problem when designing realistic models and cross sections must be extended sufficiently. This check allows the user to make sure the model does not "glass wall", i.e. water level exceeding the section data at any node. The model must have been run, preferably with the highest return period, and the maxima / minima csv file created from Tabular CSV. The program will then check each river cross-section to ensure the maximum water level is below the left and right maximum levels. If the maximum water level exceeds the maximum bank levels the program reports an error and lists the node name, which bank is exceeded and by how much. If two spills are attached to the node then this is automatically passed. If one spill is attached a check is made for corresponding levels to calculate which bank it connects to. If it cannot be matched, a "Spill Section Point Not Found Handcheck" error is reported. The user can then check the cross section does not come out of bank on the side not attached to the spill. Other errors which may occur during this check are reported alongside the node name. The node may not be listed in the csv file giving a "Max Level Not Found" error or the maximum level listed in the csv file may occur at the initial conditions causing an "Initial Condition Error" to be reported.

Floodplain Roughness

Cross sections should have realistic Manning’s roughness coefficients representing the varying elements of the land. The Manning’s n value can have a significant effect on the levels and flows produced by some models. This check examines all of the river sections in the model and looks for the outside panel markers. It then ensures that the floodplain, taken to be outside these furthermost panel markers, is rougher than the in bank channel. The program will report any nodes which fail the check with an error and state if left and / or right banks are at fault. If the panel markers do not exist then the check ignores the river section as the error would be reported in the Panel Markers check.

Manning’s Roughness in Range

Cross sections should have realistic Manning’s roughness coefficients representing the varying elements of the land. The Manning’s value can have a significant effect on the levels and flows produced by some models. A useful guide to standard Manning’s roughness values can be found in Open Channel Hydraulics (1959) by V. T. Chow. This check ensures that the roughness in each river section is within sensible limits. Users can modify the limits to suit the particular model. The check will give the following errors and warnings alongside the corresponding node name (default values are shown here, which can be changed by the user):

Each river cross-section may report multiple errors and warnings.

Neighbouring Section Roughness Rate of Change Comparison

The model should have roughness values which are graded between river-sections to avoid abrupt changes in flows. This check ensures that neighbouring river-sections have similar Manning’s values in bank. The check compares every river section’s in-bank Manning’s n roughness with upstream and downstream sections. A maximum percentage change between sections can be set by the user, or left at the default 10%. If the change is more than the set percentage change then a warning is made reporting the node label and whether the problem is the upstream and / or downstream section, along with corresponding Manning’s values.

Storage to Flow Floodplain Changes

When modelling floodplains using extended cross-sections, it is common practice to set Manning’s values to zero where storage is thought to occur. This would usually occur upstream of a structure, therefore along each reach (a section of open river or conduit separated by a junction, reservoir or hydraulic structure) there should not be an abrupt change from storage to flow, or vice versa. This check goes through each reach consisting of river sections. It will look for the lowest Manning’s value in the cross-section and compare it to the lowest values in the upstream and / or downstream cross-sections. An error is returned if one section has zero Manning’s and the other has a non - zero value. The program will output the node name of the cross-section, whether the upstream and / or downstream node is not consistent, and the corresponding lowest Manning’s values.

Node Spacing - Width, Slope, Erratic

Node spacing can be critical for an accurate and reliable model, and poor spacing can be the cause of model run failure. Sparse node spacing or erratic node spacing can cause non-convergence and affect the shape of the hydrograph thus influencing flow, storage and attenuation. The program can carry out four checks relating to node spacing and gives an indication as to where additional interpolates might be needed.

The first (Top Width Spacing) checks the initial top width against node spacing, where the node spacing should be less than the top width × 20. The program assumes that initial conditions are indicative of the normal flow conditions within the physical system. It will check river sections, conduits, interpolates and replicates. For interpolates and replicates the top width is calculated from upstream and downstream river sections / conduits. If node spacing is found to be outside the recommended range an error is reported with the node name, current distance to next and recommended maximum spacing based on top width × 20. Additionally, if an interpolate is found downstream of a conduit unit, an 'Interpolate in Conduit’ warning is output alongside the node name

The second check (Slope Spacing) looks at the average slope of each reach and checks the spacing of the nodes. The nodes should not be further than 1/(2×slope) apart. When an error is found, the program will record an error and output the node name, the current distance to next and the maximum recommended distance based on the slope. Note that artificial V’s in the channel can sometimes affect the result if the Health Check is unable to identify them.

The third check (Erratic Node Spacing) looks for sudden changes in node spacing in each reach, and will check river sections, interpolates, replicates and conduits. A percentage error, set to 50% by default, can be altered in the 'Erratic %’ box and determines the rate of change in node spacing. The program will output an error with the node names which exhibit the erratic spacing and their corresponding distance to the next values

The "Distance to next realism" check makes sure no distances to next are less than or equal to 5m. Nodes which are placed so close together can be an indicative of a poor model, where interpolates have been added purely to allow the model to run or nodes have not been equally spaced around a structure. If a node is found to have a distance to next of less than or equal to 5m an error will be recorded with the node name and distance to next listed.

Typically, single or multiple closely spaced interpolates just upstream of a structure are indicative of poor model interpolation or other numerical problems within the model. Adding extra nodes to a model does not necessarily require additional survey or be onerous.