Flood hazard maps are usually based on a unique flood extent generated by running deterministic models for a single set if boundary inputs. However, various factors can affect drainage systems and the flood hazard map generated is not generally unique. To overcome this issue, a methodology to generate stochastic flood maps with InfoWorks ICM was recently published in the Stochastic Environmental Research and Risk Assessment Journal:
This methodology considers the variable capacity of sewer inlets to generate flood hazard maps and was tested in the Zona Central catchment, Coimbra, Portugal. This catchment had suffered from several floods that have essentially been caused by the limited capacity of sewer inlets due to steep slopes and clogging during rainfall events. The catchment has an area of approximately 1.5 km2 and the ICM model includes 1008 inlets.
The methodology presented is based on running several simulations with variable capacity of the sewer inlets prone to clogging. After the identification of the critical inlets prone to clogging, their maximum capacity was considered variable according to a probabilistic distribution (Beta-PERT). Then, each run is based on a set of simulations which consider the overall results and therefore dictate a flood probability. In more detail, Monte Carlo analysis are carried out to define the number of simulations necessary for stable results. The results are based on the analysis of the maximum water depth of 2D elements, which are considered flooded if the water depth is higher than a defined threshold. In the end, each 2D element have a probability of flooding according to the set of simulations analysed.
The results presented for the Zona Central case study shows that this methodology can achieve more comprehensive flood hazard maps and can capture floods not predicted by the use of traditional deterministic models. For both design boundary inputs and for a real flood event, the generated flood hazard maps show that the sewer inlet capacity can constitute a significant source of uncertainty in urban drainage models and largely affect flooding occurrence and extent.
For the more curious and interested users of ICM, here are some details of the advanced capabilities that ICM provided for this work:
This work involved running 200 simulations for each run, so in the end more than 1000 simulations were carried out. All the work was based on a deterministic model previously set up and calibrated. The sewer inlets were modelled with orifices where the maximum capacity was considered the limiting discharge parameter. A ruby script was used to generate a set of 200 scenarios with a maximum limiting discharge for each one sampled from the Beta-PERT distribution. The ruby script is shown at in the end of this post and the next figure shows the ICM model and the list of scenarios.
Each run was set up with the 200 scenarios generated which ends up in 200 simulations as shown in the next figure.
The tool “Results > Export Maxima to GIS” was then used to batch export all the 200 simulations for each run in one process:
This was a straight-forward methodology that took advantage of the more advanced capabilities of ICM. In practice, an innovative methodology based on ICM was used to add value to a deterministic model and produce stochastic flood maps.
In summary, this is one more example of how the robust and flexible tools and functionality within InfoWorks ICM can add value to flood and urban drainage studies.
coef=[…,… ] #array with samples from the Beta-PERT distribution
puts “Scenarios created:”
scenario = “Scenario_” + i.to_s + “_” + coef[i].to_s
ro=net.row_objects(‘hw_orifice’).each do |ro|
if (ro.user_text_1 == ‘critical’)
ro[‘limiting_discharge’] = coef[i]*ro[‘limiting_discharge’]/100
puts “Well done!”
The full reference of the paper is:
Leitão, J.P., Simões, N.E., Pina, R.D., Ochoa-Rodriguez, S., Onof, C., Sá Marques, A., 2016. Stochastic evaluation of the impact of sewer inlets’ hydraulic capacity on urban pluvial flooding. Stoch. Environ. Res. Risk Assess. 1–16. doi:10.1007/s00477-016-1283-x