## Introduction

The operations functions of eWater Source have been designed around the current work practices of river operators from Australian bulk water agencies.

eWater Source allows operators to develop and assess operating scenarios efficiently utilising a modelling platform that can capture both the hydrological aspects of their systems such as storages, wetlands, link routing, groundwater interactions, and management aspects such as resource assessment, water accounting, and ownership.

eWater Source represents the hydrological and management aspects of the system as individual models/components that integrate to provide the overall system behaviour. This modular structure is very powerful as it allows operators to develop system models to the level of complexity required for their system. In the past, modelling tools used by operators have often been either too simple (eg. spreadsheets) or too complex (eg. hydrodynamic models). Information on how the individual models/components in eWater Source work is provided in the relevant sections of this SRG.

eWater Source also provides tools for loading data and advancing models as well as a unique tabular display that allows the river operator to analyse scenario results in both a temporal and spatial sense.

## Scale

As a whole, the operations functions of eWater Source operate at the river basin scale and at every model time-step. For further information on spatial and temporal scales, see the SRG sections on the various models/components available in eWater Source.

## Principal developer

eWater

## Scientific provenance

Similar types of functionality exist in industry models such as IQQM, REALM and MSM/BigMod and also within spreadsheets used for daily river operations.

## Version

Source version 4.5

## Dependencies

For information on dependencies, see the SRG sections on the various models/components available in eWater Source.

## Forecasting (river operations)

Forecast models are used to build forecast scenarios in eWater Source for inflow nodes, unaccounted differences and water user demands.

Source includes three forecasting techniques commonly used in river operations: (a) monthly average, (b) time series and (c) trend models. Modellers can also build their own forecast models using the function editor.

Forecast models can only be defined after a simulation is configured and run.

## Scale

The forecast models are applied at nodes, representing a site scale. The models are applicable at every time-step.

## Principal developer

eWater CRC.

## Scientific provenance

Similar types of functionality exist in industry spreadsheet models used for daily river operations.

## Version

Source version 3.1.0

## Dependencies

There is no dependency for Inflow nodes and Water User demands except where the models reference historical values, eg. 7 day average, in which case it is mandatory to have a warm up period that covers the model’s historical data requirements.

For unaccounted differences, observed flows must be available for the relevant gauge.

## Availability/conditions

Automatically included in eWater Source.

## Structure and processes

#### Monthly average model

eWater Source allows modellers to input monthly patterns for inflows, unaccounted differences and water user demands (orders and crop models).

The monthly model is based on making forecasts using the value from the previous time-step and trending to the average for the month expressed in megalitres per day.

A trend rate can be set between 0 and 1 by the modeller. A trend rate of 0 with a daily time-step would see the target value assigned for all days of the month; ie. the forecast flow would be independent of the antecedent flow. A target flow of 0 with a daily time-step would mean this rate is the same as recession factor, so forecast flow would be equal to antecedent flow times recession rate.

The forecast monthly model formula is:

Equation 1 |
---|

where:

*t* Time-step

*FlowTarget* Target flow rate in megalitres per day for the month in which* t* resides

*Rate* Proportion of difference between target flow rate and antecedent flow rate changed across the time-step

#### Time series

eWater Source allows the modeller to input a time series to be used for forecast inflows, unaccounted differences and water user demands (orders).

#### Trend model

Source allows users to apply forecast trend models for inflows, unaccounted differences and water user demands (orders).

The trend model is based on making forecasts using the value from the previous time-step and trending to a target value expressed in megalitres per day. The trend rate (*Rate*) must be between 0 and 1 and is set by the modeller.

The trend model formula is the same as the one for the forecast monthly model (Equation 1).

#### Time of evaluation

Monthly models, time series models and trend models are evaluated at the start of the model time-step.

If an function is used for a forecast model the modeller can set the time of evaluation.

For more information refer to the function Editor User Guide.

## Input data

The input data required is reliant on the forecast model’s parameter requirements when functions are used. If the function references model parameters other than for the node for which the forecast model applies or a value in a specific historical time period then these must be loaded/available in the project scenario.

## Parameters or settings

Table 1 summarises the parameters and settings applicable to forecasting river operations.

## Output data

The outputs can be viewed by selecting the time series from the Recording Manager or by displaying on the Tabular Editor.

###### Table 1. Forecasting parameters

Parameter | Description | Units | Default | Range |
---|---|---|---|---|

| ||||

Trend to Value | This aspect controls whether the model trends to the target value for the month or applies the target value for every day of the month | N/A | True | True or False |

Rate | This controls each time-step incremental value during the forecast period. | N/A | 0.100 | 0.001 to 0.999 |

Month | The is first column of the Monthly average table and is fixed. | N/A | N/A | January to December |

All Remaining | This aspect controls how many time-steps in the forecast period the model applies for (common to all forecast model types). | N/A | True | True or False |

Time-steps | This is the number of time-steps the forecast model is applied for (common to all forecast model types). | Model Time-steps | 1 | 1 to 99999 |

| ||||

Source | Type of time series data source | N/A | File | File; Scenario; |

Filename | Path to the file containing the time series data | N/A | N/A | N/A |

Column | This is the column in the data file that contains the time series data for this model element | Integer | 1 | 1 to the number of columns in the data source file |

Units | The units for the time series being loaded from the data source file | Various | megalitres per day | Refer Source user guide |

Reload on Run | This control must be checked if the time series data is to reload using the advance feature. | N/A | False | True or False |

| ||||

Target | This is the value the trend model is targeting | ML/day | 0 | 0 to Unlimited |

Rate | This controls each time-step incremental value during the forecast period. | N/A | 0.100 | 0.001 to 0.999 |

All Remaining | This aspect controls how many times steps in the forecast period the model applies for. | N/A | True | True or False |

Time-steps | This is the number of time-steps the forecast model is applied for. | Model Time-steps | 1 | 1 to 99999 |