Innovyze’s InfoWorks RS has been the key software in helping to formulate flood risk reduction strategies throughout the Northland Region of New Zealand, an area of approximately 14,000 km2 on the northern tip of New Zealand’s North Island.
The Northland Region has a predominately tropical climate. In recent years a number of tropical storms have affected the region, with rainfall intensities of up to 150 mm per hour, resulting in devastating flooding affecting local economy, infrastructure and life. It is anticipated that the results of Climate Change will increase the frequency, duration and size of these storms, seriously impacting the urban areas of the region.
The Northland Regional Council (NRC) identified 27 river catchments that had been significantly affected by flooding in recent years. Of these five currently had ongoing investigations.
The NRC recognised the benefits of a holistic approach to modeling an establish Manaaki Awa (River Care Group) to carry out study of the remaining 22 catchments. MWH were lead consultant in this venture and recognised that InfoWorks RS would provide the most suitable modeling software to address the client’s requirements and allow them to retain full ownership of the final models.
MWH recognised that the easy to navigate InfoWorks RS GIS interface, coupled with the accurate modeling engine would provide the best for purpose solution to the client. This is considered a progressive approach as at the time InfoWorks RS was not being extensively used in New Zealand.
The selling points of InfoWorks RS were its easy integration of the modeling into the flood forecasting and civil defence systems and the ability to easily transfer the catchment databases once complete. With further development the models can be incorporated into real-time flood warning systems to provide as much warning to communities of potential hazards.
The application of 1D, 2D and quasi 2D flood compartments was used to ensure that the most efficient runtimes were achieved from the modeling, whilst presenting accurate results.
The outputs from InfoWorks RS also integrated into the detailed GIS based Risk Management work that was carried out to identify high risk areas within each of the catchments.
The InfoWorks RS models were constructed in detailed consultation with the client. This ensured that the client was involved throughout the model build process and developed an ownership of the final models, to allow them to develop them further as catchment conditions change.
The modeling philosophy, developed by MWH, looked at a sustainable knowledge-based approach. This considered standardisation of the modeling methodology across all 22 river catchments and incorporated, where possible, existing catchment data. Where possible the NRC supplied supporting data, including rainfall and flow records, past reports and flood survey records.
The modeling approach, developed in conjunction with MWH, utilised the seamless integration of pre-processed GIS data into the model build process. InfoWorks RS was an excellent choice of software to facilitate this approach. Stability and quickness of the software allowed efficient model building to be undertaken for each catchment.
The transfer of final InfoWorks RS models will allow the client to effectively interact with the models to create future scenarios or run alternative simulations.
In addition the ease of exchanging information and results from InfoWorks RS to dedicated GIS software allowed for a variety of analysis to be carried out, including detailed flood mapping, flood risk assessment and financial implications of flooding.
Due to the anticipated over all complexity of the final models, initial modeling was carried out, based on the available LiDAR data alone. These models were used to produce rough flood maps for each catchment, to identify potential flood hot spots and areas where additional data was required to ensure model accuracy.
Where significant overland flow and flood areas were identified, but did not represent a key area of interest, the use of quasi 2D modeling was employed through the use of flood compartments. This provided a significant reduction during the detailed model runs and returned accurate results.
The model build was undertaken in a number of stages and required interaction between InfoWorks RS and dedicated GIS software. The flexibility of the InfoWorks software allowed efficient model builds to occur and an easy interchange of data between software. The database manager within InfoWorks RS allowed for the development of the models to be recorded. This creates an active audit trail through the model build process and allows the modeller to retrace to an earlier model, if required.
The initial model builds were carried out through digitising work in InfoWorks RS. Primary network centrelines and bank lines were created, with the DEM as background. Interpolation of the relevant feature and the DEM grid value allowed the bed slope of channels to be defined automatically. At the time of digitising in InfoWorks RS sample cross-section locations were identified and represented by simple lines. These were then exported to GIS for processing and definition. The sampled cross-sections were then imported directly into the InfoWorks RS model as simple shape files.
It was acknowledged early on that LiDAR data cannot successfully penetrate water surfaces. Therefore the cross-section data used in the initial models would not provide the true channel profile. It was considered that the LiDAR surface would represent the base flow component of the network. The cross-section data would need to be replaced by real data as part of the detailed analysis.
From other contour data further grids were created and these were used to define the sub-catchment boundaries required for the hydrological assessment and boundary nodes. Again simple shape files were imported directly into InfoWorks RS without the requirement of third party software.
To allow the initial models to be run and provide the rough flood maps it was necessary to create boundary nodes and connections. These boundary conditions were limited to hydrographs for the rainfall and stage data for tidal representing conditions. Connections from the boundary nodes to the primary network were carried out in InfoWorks RS through either direct connection or lateral connection along a reach to evenly distribute flows to the model. In addition to the flow and stage boundary nodes, nodes for each sub-catchment were also added, carrying the response characteristics of each sub-catchment. The catchment specific parameters, including runoff coefficients, impervious values, rainfall profiles, were determined through assessment of the aerial images and through external assessment. These were kept as a standard for all the Priority Catchments to ensure an easy to understand model for the client.
For Priority 1 Rivers it was necessary to create four hydrologic models considering the 10, 50, 100 and 100 plus Climate Change year ARI models. This allowed for a greater understanding of how flooding in the catchment develops with increases in intensity. For Priority 2, 3 and 4 Rivers it was considered only necessary to consider the 10 and 100 plus Climate Change ARI model.
Additional data was obtained from site, including specific cross-section data of hot spots identified in the flood mapping and at hydraulic structures that would impact on the flow regime. This data was added to the initial InfoWorks RS model and a full 1D analysis was carried out, to ensure stability of the system and identify any critical areas that required more survey work.
For Priority 2, 3 and 4 Rivers, once the 1D model was stable within InfoWorks RS, final design runs were made and verified against historical data. Once a satisfactory fit had been made it was then a case of producing flood maps and shape files of the resultant flood depths.
Priority 1 rivers required significantly more detail to ensure that accurate calibration and verification. To ensure accurate flow routing through the primary network cross-section data was needed at relatively short intervals. Obtaining cross-section data at the required level of detail was not feasible. To work around this the interpolation tool within InfoWorks RS was used. Surveyed cross-sections were recorded at 500 – 1,000m spacing and these were imported direct into the InfoWorks RS models as simple xyz files. The interpolation tool was then used to create cross-sections at 5, 10 and 20m spacing to provide sufficient detail.
Calibration of the Priority 1 Rivers initially considered flow balance analysis alone, to ensure that parameters and assumptions made during the hydrological modeling provided accurate results. When flow balance was achieved variations in channel roughness values was undertaken to match the observed peak data to the modelled data. This was particularly challenging but the ease of amending relevant values in the InfoWorks software aided in speeding up the process.
Once Priority 1 Rivers were calibrated it was possible to verify the models against historical and observed data. As with the calibration, flow balance and coincidence of peaks were assessed using channel roughness’s applied earlier. Once this was satisfactory it was possible to produce the flood maps and flood depth shape files for further analysis.
Potential Future Uses of InfoWorks RS River Models
Although the primary role of the InfoWorks RS river models is for flood mapping within the Priority River catchments, the potential future applications for the model, by the client are huge. The flexibility of the InfoWorks software, to interchange data with GIS and third party software means that the client has a powerful tool to work with.
Potential future applications being considered by the client include prioritising works within catchments and incorporation of the models into the civil defence scheme to provide flood warning systems.
The InfoWorks models could also be used in planning applications and land use changes that occur within the catchment over time. The user interface allows easy alteration of all elements of the model to allow the models to be kept up to date.
MWH recognised the benefits of the InfoWorks RS software for application on the Northland Region Priority Rivers project. The easy to navigate GIS interface, the ability to exchange GIS data with external software and the accurate modeling engine made this an easy choice.
The Northland Region Priority Rivers Project is the first time that InfoWorks RS has been used at this scale in New Zealand.
Innovative transfer protocols in InfoWorks RS results in the client being delivered a single master database. This ensures easy transference of the models with no risk of data becoming detached or lost in the process. This was particularly important for this project where MWH worked with another consultant modeling the catchments.
One of the most beneficial elements of the InfoWorks software is the version tracking tool. This allows backups to occur and separates them from the working model. This allows an active tracking of model development to occur and ensures that any model can be activated to work on at any time. This is particularly useful when working on large catchments and when making significant alterations to the model. The tracking tool allows the user to retrace their steps in a methodical manner.
The methodology developed in conjunction with MWH and the application of the InfoWorks RS software has resulted in a real sense of ownership of the model, by the client. This will ensure that the models are referred to in future and amended to reflect changes in the catchments. Continued support from Innovyze will also ensure that a degree of future proofing will be provided to the client.