MUSIC Frequently asked Questions (FAQs)

The trial version has a few limitations, as listed below:

- Only 5 nodes per model

- No Access to MUSIC-link

- You cannot save the model

- No import or export of files

- You cannot build templates of your own; you have to use a meteorological template pre-loaded in MUSIC ( (6hrly one year data for major cities in Australia is available))

MUSICX has been re-designed and re-written into a modern software coding platform, maintaining all the capabilities of MUSIC V6.3 but giving users additional functionality and the benefits of modern software architecture. In the current release, MUSICX allows you to combine your MUSIC models with eWater Source (and the new Source plugin - Urban Developer) to model water quality and quantity.  A limitation is that MUSICX and Source will have to run at the same time step. We are working on providing a solution where the three models will be able to run at different time steps appropriate to the scale of the different tools. This will ultimately give you the tools to design integrated water resource management solutions.  The new coding platform brings the ability to customize your MUSICX models by developing plugins. Plugins can be written for a wide range of purposes, including incorporating or testing new treatment options, adding new contaminants, recognizing local features, such as soil type and hydrology, or including new science. Compatibility with the Source platform allows you to access Source plugins, such as additional Rainfall-Runoff models. MUSICX also brings the benefits of:

- Improved Graphic User Interface for easy data input

- Tabular input/output, group edit, and advanced graphics.

- Function Manager to support more complex model building.

- Powerful in-built Results Manager for data analysis and results interpretation.

A primary driver for developing MUSICX was to support the integration of MUSIC functionality with eWater Source (including the Urban Developer Plugin) to allow for a holistic assessment of water quality and quantity water, giving water modellers and managers the tools to design truly integrated water management solutions. MUSIC and Urban Developer were designed to complement each other and were able to draw together information on water quantity (Urban Developer) and water quality (MUSIC) to develop IWRM solutions in urban areas. These models can now be linked to Source catchment and river system models. Giving water managers tools to explore possible interactions and to identify and test IWRM options across catchment and local scales, with a range of land-uses. Users can customise MUSICX by creating plugins. Plugins can currently be created to for data processing, input sets, exporting and reporting functions.

This question answers several questions around new functionality and incorporating new technology or science, such as the type of contaminants, sediment and nutrient models, rainfall-runoff models, treatment methods and new science. MUSICX has the same underlying science and capability as MUSIC 6.3, including types of contaminants, treatment nodes, source nodes, rainfall-runoff and infiltration models. eWater is a not-for-profit, Government-owned company, tasked with the ongoing development and adoption of Australia's hydrology modelling tools. All our revenue, from licenses, support and maintenance agreements and project work are re-invested into maintaining and improving our tools. Over time, we envision that together with the MUSIC user community we will add new science and capability through the creation of plugins. Similar to our approach for developing the Source platform, we will work with the MUSIC community to identify priority areas and opportunities to co-invest in projects that support the ongoing development of MUSICX. Potential areas for plugin development include: new contaminants, customized pollutant generation algorithms, new treatment types, such as a separate treatment pit, reflect the seasonality of nutrient treatment and additional rainfall-runoff models

We recognize the importance of gaining widespread acceptance from the MUSIC community on how to best include new features. Plugins will allow the community to test new approaches and treatments, from which we can work together to determine the best approach for representing new science and capability in MUSICX.

MUSICX has been designed to allow users to incorporate plugins, as described above plugins allow you to incorporate local conditions, such as soil characteristics and hydrology.

eWater provides training in writing plugins. To ensure that new plugins meet the required standards, we require all those who want to write plugins to complete this training.

This answers several questions related to compatibility. eWater established a Technical Panel of MUSIC experts to guide the development of MUSICX. In collaboration with the Technical Panel we undertook extensive testing to ensure that results from MUSIC 6.3 are reproducible in MUSICX, this included regression testing by eWater's hydrologists and members of the Expert Panel set up to advise on the development of MUSICX.MUSIC 6.3 files (.msf) can be imported to MUSICX. When you import MUSIC 6.3 to MUSICX you will need to configure climate data. MUSICX accepts a variety of data formats including; .mlb (MUSIC meteorological template file), .csv, .txt and .dat. MUSIC 6.3 cannot import MUSICX files. MUSICX is a major change to the software. We know that some of our community need time to learn the new features and transition to MUSICX. To ease the transition, MUSIC 6.3 remains available for use and eWater will continue to provide support services.

This can be done through Feature Table.

Yes, MUSICX has the similar capability as MUSIC 6.3.

MUSICX can be used in any soil conditions, you just need to change the default soil properties to reflect your needs.

MUSICX supports various data formats, including csv, .txt and .dat. You need to define the data for rainfall/evapotranspiration.

This bug has been fixed.

Access to the Pluviograph data tool is provided to user accounts which have a MUSIC software licence and an active maintenance and support subscription (SMA).

In most cases there are options for eWater to provide students with special access to products. These requests need to be assessed on a case-by-case basis by contacting our support desk. When submitting your request be sure to provide the following:

- A basic overview/abstract of the type of research being undertaken with the software; and

- Details of your study including the type of study, institution and contact details for your supervisor.

eWater will review the information provided and contact you with options for access.

We need to better understand the geometry of the building and how it is sited relative to prevailing weather in order to comment on whether the facade should be included in the catchment. In many cases, vertical surfaces have no net impact as they can also provide shelter from rainfall, depending on the wind direction.

The soil type influences the shape of the unsaturated flow curve and effectively replaces the particle size, though its function is different. For coarser and more freely draining soils, the curve is steeper which means the hydraulic conductivity of the soil decreases more rapidly as it desaturates. Therefore, changing this will influence the pattern of desaturation of the system. Usually the differences are not huge, but remember that it affects the soil moisture, and low soil moistures lead to poor nitrogen retention. In extreme cases, oversizing can reduce performance due to excessive drying. The important thing is to choose the soil type that most closely represents the real system. Modifying the soil properties such as TN Content, Organic Material, Orthophosphate and vegetation will result in significant differences in treatment performance and should be chosen with care to be representative and effective. See the FAWB guidelines for more guidance but generally lower is better.

In the music v4 bioretention node, Filter Depth is the depth of the filter medium alone, Submerged Zone Depth is the depth of the submerged zone alone, and the total depth of fill material in the system is the sum of the two. The filter medium is fine enough to support vegetation and to control the flow rate through the system. The submerged zone, by contrast, assumes material that is coarse enough not to restrict the flow. When a drain pipe is present (as it usually is), it is located at the junction of the two zones.

Sure, it would be best to use the infiltration node. The node represents an infiltration trench below the surface with a ponded zone above the surface. MUSIC represents infiltration vertically from the base and also horizontally through the sides of the system. MUSIC does not represent clogging so it is important with infiltration systems to ensure that effective sediment treatment is provided and it is also prudent to adopt a conservative exfiltration rate. Water should be treated to a suitable standard for groundwater prior to entering the infiltration trench, particularly in freely draining soil conditions. 

In MUSIC v4, the bioretention node algorithm has been significantly modified. The filter media is now explicitly represented as a soil moisture store, and the water quality algorithms now include consideration of the soil pollutant concentration, vegetation, submerged zone and the soil moisture prior to commencement of an event. In v3, the outflow pollutant concentration was a simpler function of inflow concentration and particle size (eg a coarser filter media such as sand has poorer treatment performance). This has now been replaced with the considerations above.

Absolutely! Use an urban node set to 100% impervious to represent the roof, and a bioretention or media filter node to represent the raingarden. If the raingarden is vegetated and uses a filter medium something like a sandy loam, the bioretention node will be more appropriate. If the raingarden is really more like a sand or gravel filter, then use the media filter node. If you also want to model urban areas other than roofs, use another urban node for these areas, and bring it in downstream of the bioretention or media filter node. You will need to adjust the fraction impervious to represent the remaining urban area with the roof missing. You can then lead the combined outflow on to the catchment outlet or to any other treatment nodes that you have in mind. If you want to model many roofs each with its own raingarden, and they are all the same, you can lump them up to one big roof and one big raingarden. Just add up their respective areas, and don't change any depths. If each combination is different - big roofs with small raingardens and vice versa - the lumping will not be accurate. There are sure to be some ineffective combinations among them. In this case you really need to model each one separately.

It is good modelling practice to parameterise your model specific to the location and conditions you are modelling. Paramaterisation of a model includes the selecting the correct rainfall data, defining the catchment and ensure that the pollutant characteristics are representative of your catchment. MUSIC provide default parameters which in some cases will be appropriate though in all cases it is recommended they are reviewed and changed as needed. Generally you should adjust pollutant loadings to differentiate between roofs and ground level impervious surfaces. It is recommended that you seek guidance from your local authority or regulator to ensure you are setting up your model according to their requirements. Refer to Australian Runoff Quality - A guide to Water Sensitive Urban Design for general guidance on pollutants by land use category or Urban Stormwater Quality: A Statistical Overview by Hugh Duncan - CRC for Catchment Hydrology). Additional resources include:

- Water by Design MUSIC Modelling Guidelines

- Using MUSIC in Sydney's Drinking Water Catchments

- Melbourne Water MUSIC Guideline

The stochastic generation of stormwater pollutants will create a new set of pollutant concentrations each time a source node is run (which only occurs when it is modified or the model is closed and reopened). This means that each time the model is run, the results for pollutant load reductions will be slightly different. However, as long as there are a large number of independent numbers in the data set, the results should approach a mean with variations of less than perhaps 2-5%. If you run the model for one year of daily data, you will have 365 predicted concentrations. If you use 6 minute data, you will have 86400. Obviously, the longer the data set, and the smaller the timestep, the more individual data points you will have. This will help to smooth the variability of results between separate runs. If you run just one year of data or a daily timestep, the variations in results can be significant because the data set is too small. 

A second issue is that serial correlation is enabled. This means that the concentration at one timestep is related to the concentration at the previous. Pollutant concentrations vary with time and there can be significant fluctuations, which make the use of a mean concentration unrealistic. However, there is also good correlation between consecutive time steps and the serial correlation allows this to be represented and provide more visually reasonable concentration time series. A side effect of this is that because concentrations are now related, it effectively reduces the number of independent concentration measurements, because all concentrations within a given event will be correlated. As a result, more variability is observed. Given this consequence, you may prefer to set the serial correlation to 0. This will still provide a good estimate of performance, better than using a mean concentration (which will over-estimate performance), but reduce the variability of your results. I would also recommend that you run the model on a minimum of 5 years of 6 minute data to minimise variations in results and most contemporary guidelines require this. If there is any doubt in the results, do several runs and take an average. 

The baseflow component of runoff is harder to model than the stormflow component, because there are more parameters to calibrate, but you should be able to get an approximate fit provided the basic assumptions are satisfied. First of all, do your observations balance? Can the catchment physically produce the larger volume of baseflow you are observing from the measured rainfall? If not, there could be other inputs from groundwater movement or garden watering runoff or water supply system leakage. If it does seem to balance, what are the long term proportions you are trying to model? How much rainfall on average becomes stormflow, baseflow, and losses? Then it comes down to tweaking each part of the model separately to achieve these proportions. All the parameters interact to a greater or lesser extent, but only the impervious rainfall threshold, the deep seepage rate, and the ET file can directly affect losses. The fraction impervious and the impervious rainfall threshold are the most important factors for stormflow, but the two infiltration parameters can also have an effect. Baseflow is affected by more parameters, so try to get a first estimate of losses and stormflow first. When the long term baseflow proportions are reasonable, you can modify the speed of response using the groundwater depth and the daily baseflow rate. Lower initial depth and higher baseflow rate gives a sharper response, while higher initial depth and lower baseflow rate gives a slower response. Keep in mind that all the factors in Music are intended to be calibration parameters, not immutable physical constraints, so any reasonable value that successfully models observed flows is a good value to use. Fraction impervious is probably the most important single parameter in Music. It seems easy enough to measure total impervious area in a catchment, but it is not nearly so easy to measure effective or directly connected impervious area, and that is what is required. So start with the value that gives the best calibration. If it is very different from what you expected you will need to find out why, but it is still a good value to use for modelling. Similarly for all the other parameters.

In almost all cases the inflow will ONLY be the impervious area runoff (ie. not the baseflow).  Why?  Because the bioretention system will be receiving water from either direct runoff from an impervious surface (e.g. streetscape) or receiving water from a pipe. In either case, baseflows water infiltrated into the ground and then coming back out into the waterway) will not pass through the treatment node.The best modelling option in MUSIC to take into account this baseflow is to use a secondary drainage link to transfer the baseflow directly from the source node the the downstream point. Volumes will usually not be significant though and even this solution may be discussed since the transfer time of the baseflow water from the source to the receiving waterway is still a topic of active research. 

MUSIC allows the user to define your rainwater tanks properties individually and then MUSIC generates the lumped total values. Nevertheless, if you want to set up a water demand out of those tanks, the volume you entered into the re-use box has to be the total reuse volume, summing up the reuse of every tank. For example, if the assessing authority requires a re-use of 3000 kL a year per tank, and there is 5 tanks in the development, then the volume you entered in the node has to be 5 tanks * 3000 kL/tank/yr = 15000kL /yr.