WRM web version: Reaches and the Reach Editor

Several of the nodes, and the reaches connecting them, can also be linked to one or more underliying aquifer to describe conjunctive use and interactions of surface and groundwater.

A reach connects two nodes hydraulically; it represents unsteady open channel flow in natural or man made channels. Reaches are assumed to be hydraulically homogeneous, their length should be less than the flow travel distance within one time step (i.e., reach length shoud nor exceed between 20 and a maximum of 40 km). If a length is too long or changes its characteristics without any structural (confluence, diversion) element to break it, a geometry or auxiliary node can be used to start a new reach with new geometric and hydraulic properties.

Reaches are abstract constructs: while they may represent real sections of natural or man made channels, they can also be of zero length to facilitate the combination of several nodes at the same location to obtain more complex behaviour.

The reaches are used to route the unsteady open channel flow using the Muskingum method; resistance, slope, length of a reach as well as a weight factor to describe the respective importance of inflow and storage within the reach are used.

Reach editor

Once a reach is selected, the corresponding editor page shows its location in the network graph, and its properties that can be edited:

Reach properties

1. Name (short but descriptive) and ID (internal);
2. Connectivity: start (head) and and end (tail) node (from - to);
3. Length, in m (distance between the nodes connected);
4. Slope this is computed from the elevation difference of start (head) and end (tail) node; please note that this elevation difference must be >= 0. i.e., the start node must be above or at the same level as the end node of a reach - this is being checked by the plausibility part of the consistency checker.
5. Resistance or Roughness, Mannings n used to estimate flow velocity; in the editor, we do NOT ask for roughness explicitly, but offer a range of channel types from which an appropriate coefficient is derived internally; examples for field tested values are:

   Channel type/surface	M_n	Channel type/surface	M_n
   Natural stream, smooth, straight	0.025	Excavated channels, smooth	0.022
   Natural stream, straight, cobbles/gravel	0.030	Excavated channels, gravel	0.025
   Natural stream, meandering	0.035	Excavated channels, vegetated	0.030
   Natural stream, sluggish, pools	0.040	Excavated channels, cobbles, boulders	0.035
   Floddplain, pastures	0.035	Lined channels, finished concrete	0.012
   Floddplain, brushland	0.050	Lined channels, unfinished concrete	0.014
   Floddplain, heavy brush, forest	0.150	Lined channels, PE, PVC	0.010

6. Weighting factor (for the routing); this describes the relative importance of inflow versus storage for determining the outflow; default value is 0.2; values would typically range from 0 (no direct effect of inflow) to 0.5 (inflow and storage contribute to outflow equally) to 1.0 (only the inflow controls the outflow, no storage effects);
7. Cross-section: for the default trapezoidal shape, this consists of values estimated at mean flow: .

   depth (in m)

   upper (surface) width (in m)

   lower (bottom) width (in m)

   reference flow (in m³/s

   where reference flow is the flow (m³/s) corresponding to the depth specified with the trapezoid channel geometry approximation.

8. Reach Water Budget: The water budget of a reach is primarily defined by
   • its inflow (from its head node),
   • its outlow (at the tail node),
   • the storage change in the reach itself.
   
   Additional (optional) processes that can be represented include:
   1. Direct precipitation and evaporation;
   2. Lateral inflow (surface runoff and interflow from the unsaturated zone);
   3. Groundwater interaction.

   Direct precipitation and evaporation are estimated in analogy to the corresponding processes for a reservoir:

   reach surface area (defined by reach length and the surface area from the rating curve, see below) are used together with an optional local temperature and precipitation time series - if these are not specified for a given reach, the scenario level default data sets are used. A local degree-day coefficient for evapotranspiration can be used to adjust for water temperatur and bank vegetation.

   Lateral Inflow, represents the storm runoff contribution (surface runoffand interflow) of the immediate reach catchment, not represented by a subcatchment start node and a channel with an explicit confluence.
   Data requirements include the specification of the local catchment area together with a runoff-coefficient that defines the fraction (in %) of precipitation (weighted by the catchment area) that will reach the channel as surface runoff or interflow in a given time step (no delays on a daily scale).

   Groundwater Interation: or linkage to an Aquifer: the reach can interact with an aquifer for infiltration and exfiltration (groundwater or saturated zone contribution to the lateral inflow or baseflow contribution, and seepage into the groundwater from the river channel); the (optional) data requirements include:

   • Loss coefficient: a simple % seepage loss of a fraction of a reaches flow to the aquifer coupled to the reach.
   • Minimum flow: a minimum flow value, below which no water will seep to the aquifer;
     SEEPAGE = MAX (0,(FLOW - MINIMUM FLOW )) * LOSS COEFFICIENT
   • Initial distance, im m, of the water level to the groundwater head, corresponding to the initial flow; by default, this is assumed to be in equilibrium at the start of a simulation, that is, the initial flow and corresponding river and groundwater levels will cause neither infiltration nor exfiltration: these will be driven by subsequent changes in the respective levels.
   • a Coefficient describing the infiltration/exfiltration behavior (hydraulic conductivity) of the reach vis a vis the groundwater. This is estimated as follows:

      FILTRATION_FLOW = DELTAh * WP * COEFF 

     where DELTAH (in m) is the elevation difference betwen the aquifer and the average surface level of the reach, WP is the wetted perimeter, (obtained from the rating curve/cross section data for a reach) and COEFF is in m2/s .

Speed of open channel flow

The simple estimator used is:

V = 1/M * Rh**0.67 * SQRT(SLOPE)

In terms of frictional head losses, the perimeter is important. Hydraulic radius, Rh,

Rh = A/Pw

Reach Geometry

1. minimal requirement: reach length and average depth and width (at mean flow);
2. optional extension: the rating curve (see below) relating depth/level with flow;
3. optional extension: a complete geometry TABLE, relating:
   • flow in m³/s
   • level or depth (in m)
   • surface width (in m)
   • wetted perimeter (in m)

Rating Curves

A RATING TABLE or RATING CURVE defines the relationship between flow (in m3/s) and a depth reading for the segment - while normally used to derive flow data from level monitoring, in WRM they are used to obtain estimates of LEVEL from the flow calculated by the model. Together with the DEM and a georeferencing of reaches and their Cross-Sections, this can be used for the prediction and monitoring of flood conditions.

The values can either be tabulated as pairs of L (level or stage) and Q (flow, discharge pairs), or expressed by an exponential curve of the form:

Q = aL**b

Quoting the NOAA definition:

A rating table or curve is a relationship between stage and discharge at a cross section of a river. In most cases, data from stream gages are collected as stage data. In order to model the streams and rivers, the data needs to be expressed as stream flow using rating tables. Conversely, the output from a hydrologic model is a flow, which can then be expressed as stage for dissemination to the public.