Rainfall/Evaporation Boundary

Data

Theory and Guidance

The rainfall evaporation boundary unit (REBDY) provides a rainfall and/or evaporation boundary inflow into a model network. This is of particular interest where direct rainfall or evaporation forms a significant proportion of the water volume entering a system. There may be up to three consecutive tables of data, representing rainfall, evaporation/evapotranspiration and/or infiltration. There is the opportunity to use actual dates or standard time series (as in the QT-type boundaries) with the rainfall data and flow averaging will be done within the REBDY unit so values can be used directly in mass equations.

The REBDY unit always operates in conjunction with either a lateral inflow unit, or a lateral inflow node of a RIVER or RESERVOIR unit, to create an inflow into one or more units with a magnitude based on the water surface area of the receiving unit. A REBDY unit must either be connected to exactly one lateral inflow unit or to any number of lateral inflow nodes of a RESERVOIR or RIVER unit. This can then divide the resultant inflow across the network accordingly. The lateral inflow unit must have its distribution method set to AREA for the REBDY unit to apply correctly.

The REBDY unit can hold combinations of rainfall data, evapo[transpi]ration data and infiltration data. These are input via individual tabs on the REBDY form within the Flood Modeller interface. It is therefore, for example, possible to store rainfall data with a different time interval from corresponding evaporation data. If both rainfall and evaporation are present then the REBDY unit will combine these into an effective rainfall figure during the computational phase of the model, i.e. (rainfall - evaporation). Similarly infiltration will be treated as a negative value in the calculation of an effective inflow.

The corresponding time data for rainfall and evaporation data can be in any of the formats already available for QT-type boundaries, i.e. seconds, minutes, hours, weeks, etc. In addition, any of the available time units may be specified in a date format. The date format within the data form must be of the form dd/mm/yyyy hh:mm. It is recommended that the user choose the Select ('...') button within the data grid to bring up the following dialogue which ensures the correct formatting.

The typical method of specifying rainfall and evaporation data is as a depth intensity (Data format = INTENSITY) with units of depth per unit time, which can be considered as flow per unit surface area. The REBDY is designed by default to expect this form of unit, however, a variety of intensities are available, e.g. mm/15 minute, mm/hr, mm/day, etc. or even a user specified intensity time interval (in seconds). These may be specified by altering the Intensity Time Units field. An aggregated rainfall depth over the preceding data interval may alternatively be supplied (Data Format=DEPTH), in which case an intensity is obtained by dividing the depth by the preceding time interval. In order to convert the intensity value to an equivalent inflow over a model time step the REBDY unit calculates the average value over the time step and this is then multiplied by the receiving surface area in the lateral inflow unit to obtain an inflow.

Example

To explain how the averaging of date format rainfall data will be done; consider a typical rainfall distribution, as shown below. The rainfall time interval is different from the model time step, dt. As the shaded area shows there are two different rainfall rates occurring over the second model time step. The total depth of rain falling during the time step is equivalent to the shaded area, V. Dividing this value by the model time step, dt, gives the average rainfall rate per unit time, with units ms-1. Finally, multiplying this value by the surface area receiving the rainfall will produce an equivalent inflow. This is therefore dependent on the dimensions of the receiving river or reservoir unit.

The REBDY data can have irregular increments between data points in a series as the unit will always calculate the time between data points. For bar chart format data, as shown below, for any model time preceding that of the first data item, the first data value is assumed to apply up until this time. NB Bar chart format is forced if data is specified as a DEPTH (as opposed to an INTENSITY). In addition, a repeating or extended series can be specified, which may be useful for example when specifying an annual or daily evaporation pattern (see Notes).

Notes

1. A rainfall, etc. value corresponds to the intensity or depth that occurred during the data interval prior to its corresponding time. If the model start time is before that of the first data item, this value is assumed to occur up to the latter time.
2. Since the first value in the data series applies up to and including (but not after) its corresponding time, using the REPEAT methodology will effectively repeat the data from the second data value (i.e. that applies between the first and second specified times). It may therefore be advisable to insert a dummy item at t=0.
3. Using REPEAT will never repeat REBDY data 'backwards' in time - the first data item will always be extended backwards in such cases.
4. Any Hydrological boundary (e.q. ReFHBDY, GERRBDY, FEHBDY, -etc) may be treated as an REBDY by selecting the 'Hyetograph' option on the Options tab within its unit form. This will then apply the generated hyetograph to act as a direct Rainfall boundary.
5. If DATES are entered as time units, then the model run times (either the Start Time or Time Zero) must be specified as a Date and Time.
6. If an REBDY unit is attached (via a lateral inflow unit) to a reservoir, there must be at least one regular inflow/outflow node (i.e. non-lateral inflow) also attached to the reservoir. If none is naturally attached, then a dummy node (e.g. a QTBDY inflow with zero flow, or a weir with an unfeasibly high crest) must be attached.

Errors

The following errors may be generated:

Datafile Format

Line 1 - keyword "REBDY", version number "#REVISION#2" , [comment]

Line 2 - Label 1

Line 3 - rtype(1), ... rtype(ntyp)

The following block is repeated for n = 1 to ntyp, where ntyp is the number of types of specified data (maximum 3)

Line 4 - ndat(n) [tlag(n)] [z(n)] [tm(n)] [repeat(n)] [smooth(n)] [qmult(n)]

Line 5 - Intenstr(n), diflag(n)

Line 6 to 5+ndat(n) - depth(n,i), time(n,i)[, timed(n,i)]

Data File example