Model Setup

This tutorial focuses on how to develop a 2D flood model using an existing calibrated 1D stormwater model. The base model, provided in the “Tutorial_1D.gdb”, represents an urban watershed that includes both pipe networks and open channel systems. We will build upon this 1D model to generate a 2D flood simulation.

To begin, open the “Tutorial_1D.gdb” project in GeoSWMM. Then, add additional spatial data layers to start 2D modeling.

Required Data Inputs:

  • DEM (Digital Elevation Model)
  • 2D Model extent Boundary
  • Obstructions (Building footprints)

All of the above datasets are included in the ‘Supporting_Data.gdb’ file provided in the tutorial folder. Once all these layers are loaded to your ArcGIS Pro workspace along with the GeoSWMM project, follow the steps below to build the 2D model from the existing 1D GeoSWMM project.

Step 1: Enable 2D Modeling

To begin, click on Enable 2D Modeling. This activates the tools required to develop the 2D flood model based on the existing 1D stormwater model.

Step 2: Define the 2D Boundary Layer

The boundary layer determines the spatial extent of the 2D model. To minimize excessive computational load, the boundary should be focused around areas of interest-typically along the stormwater conduits where surface overflow is expected.

This tutorial includes a pre-defined boundary layer. The user may also create a custom boundary layer using the instructions in the GeoSWMM User Manual. The attributes assigned to this layer are as follows in Table 2.2:

Table 2.2 : Boundary Layer Attributes

Resolution

Cell Type

Roughness

Seepage

Remarks

20

Hexagonal

0.03

0

Floodplain

5

Hexagonal

0.03

0

Open Channel


Step 3: Create 2D Nodes

Using the DEM and Obstructions layers, generate the 2D nodes by running the Create 2D Nodes tool. This tool automatically places nodes across the model based on the resolution defined in the 2D boundary layer.

If the resolution is too coarse (missing key topographic detail) or too fine (resulting to long simulation time), adjust the boundary resolution settings and regenerate the nodes. Once satisfied with the grid density and coverage, proceed to build the 2D model layers.

Step 4: Develop 2D Layers

After generating the 2D nodes, use the Create 2D Layers tool to build the full set of 2D hydraulic layers that form the 2D overland flow grid. The Nodes_2D layer provides elevation data for developing these layers. GeoSWMM 2D automatically deactivates any open channels that overlap with the 2D overland flow grid to avoid double counting flow paths or spatial coverage. At this point, the overland model is in place and ready to be connected with the underlying stormwater network.

Step 5: Connect 1D and 2D

To realistically simulate urban flood dynamics, the 1D pipe network must be hydraulically connected with the 2D overland flow grid. In this tutorial, Orifice elements are used to establish these connections. They serve as hydraulic control points that enable bidirectional flow between the overland and the subsurface systems—allowing the model to capture surcharge, overflow, and ponding behavior with greater accuracy.

Step 6: Define Boundary Conditions - Add an Outfall for Surface Drainage

In 2D overland flow modeling, surface water can accumulate near the boundary of the study area —especially near the system 1D outfall—if no additional outfall is assigned within the 2D grid. While some localized inundations near these boundaries may be expected, foregoing 2D outfall nodes altogether can result in exaggerated or unrealistic flooding, particularly at low-lying edges of the study area.

To ensure accurate simulation of overland drainage, it is recommended to add 2D outfall nodes at appropriate locations within the model domain. These newly added outfall nodes serve as surface drainage exit points, allowing water to flow continuously and naturally out of the system. This maintains hydraulic continuity and prevents artificial ponding along the 2D model boundary.

Tip: Assign 2D outfalls at the lowest elevation points where overland flow would naturally exit the modeled area. Refer to Figure 2.2 below for a visual example.

T10_Figure 2.2 : 2D flow grid before (left) and after (right) assigning an outfall at the lowest elevation.png
Figure 2.2 : 2D flow grid before (left) and after (right) assigning an outfall at the lowest elevation

For this tutorial, an outfall node OF2 was added at the lowest point in the 2D surface grid near the study area outfall. This ensures realistic surface drainage and more accurately reflects real-world flood behavior.

Rainfall Data

In GeoSWMM, a rain gage is used to provide precipitation or rainfall input to the model. For this tutorial, the study area is assigned a 24-hour synthetic storm event with three return periods: 2-year, 10-year, and 100-year. The rainfall data is linked to the rain gage through the Rain Gage Property Editor, as shown in Figure 2.3.

T10_Figure 2.3 : Rain Gage Property Editor.png
Figure 2.3 : Rain Gage Property Editor

In this tutorial, the data source is set to TIMESERIES, and the series name is “100yr24hr”. This time series is sorted and managed within the Time Series block of the Model Object Panel in GeoSWMM.

For detailed instructions on assigning rainfall data and creating time series, refer to the GeoSWMM User Manual.

Time Series Data

The Tutorial_2D.gdb project includes a predefined time series dataset located under the Time Series block in the GeoSWMM model panel. This dataset represents a 100-year return period, 24-hour duration synthetic storm, and is directly linked to the Rain Gage property. The time series is formatted using intensity values with a 6-minute time interval (Figure 2.4), providing high resolution for simulating detailed rainfall-runoff responses throughout the model domain.

T10_Figure 2.4 : Design Storm Hyetographs.png
Figure 2.4 : Design Storm Hyetographs

Simulation Options Set Up

Simulation settings are configured in the Options block of the GeosWMM Model Object Panel. The Options editor contains five tabs, each allowing control over different aspects of the simulation.

For this tutorial, a 24-hour long simulation has been carried out at 1 minute reporting time step to support overland flow analysis. Key simulation parameters, including process model, routing methods, and time step controls, are summarized in Table 2.3.

Table 2.3 : Simulation Options for Tutorial 10

Parameter

Setting

Remarks

General tab

Process Models

(activated and checked)

Rainfall/Runoff

Flow Routing

Input and analysis type

Infiltration Model

Green-Ampt

Physically based infiltration method

Routing Model

Dynamic Wave

 

Used for full hydraulic routing in the conveyance system

Allow ponding

Checked

Enables water to temporarily accumulate at nodes

Minimum conduit slope (%)

0

Allows flat conduits

Dates tab

Start Analysis on

1/1/2024   00:00

Simulation start time

Start Reporting on

1/1/2024   00:00

Matches analysis start time

End Analysis on

1/2/2024   00:00

Defines a 24-hr simulation

Time Steps tab

Reporting

00:01:00

Time interval for result output

Runoff: Dry Weather

00:00:05

Time interval for dry conditions

Runoff: Wet Weather

00:00:20

Time interval for wet conditions

Routing

0.5 Seconds

Computational time step for hydraulic routing

Dynamic Wave

 

 

Inertial term

Ignore

Simplifies momentum equation

Normal flow criterion

Slope and Froude

Governs switching between flow regimes

Force main equation

Hazen-Williams

For pressurized flow in force mains

Surcharge method

Extran

Handles surcharged conduits conditions

Use variable time steps, adjusted by

75%

Adaptive time stepping based on system conditions

Minimum variable time step

0.001

Sets lower bound for adaptive stepping

Time step for conduit lengthening

0

No conduit lengthening applied

Minimum nodal surface area

1

Used for ponded volume calculation

Maximum trials per time step

20

Max iterations for hydraulic solution per step

Head convergence tolerance

0.005

Tolerance threshold for flow stability

Number of Threads

1

Single-thread processing (can be adjusted for performance)

NB: All other tabs and parameters are left at their default settings for this tutorial.

Loss Parameters

In hydrologic modeling, two major forms of water losses are infiltration and evapotranspiration. To account for infiltration losses from the subcatchments, Green-Ampt model has been applied in this Tutorial. Uniform parameter values have been assigned across all subcatchments, as shown in Table 2.4.

Table 2.4 : Green-Ampt Model ParametersTable 2.4: Green-Ampt Model Parameters

Parameter

Value

Unit

Suction Head

7

inch/hour

Conductivity

0.01

inch/hour

Initial Deficit

0.01

1/hours

A monthly average evaporation rate of 0.01 in/hr has been applied to the model. For simplicity, depression storage is assumed to be uniform across all the subcatchments. This approach ensures consistent surface runoff behavior, as overland flow is only generated once depression storage is satisfied.