Explore flooding in a watershed

Flood extent and depth vary depending on the terrain and the intensity and duration of rainfall. The frequency of different flood levels depends on the frequency of the corresponding rainfall events. Where roads cross streams, floods can pose a severe hazard to drivers if the water flows over the road, which can happen when the stream channel is conducted under the road in one or more pipes or tunnels, known as culverts.

In this tutorial, you'll use Arc Hydro tools in ArcGIS Pro to model the depth and frequency of flooding in the Bear Creek watershed and model the flow levels where road crossings in the watershed become hazardous due to the flood water. It is important to be able to model these conditions for emergency management, to close roads when they are likely to become unsafe, and for prioritizing road crossings for infrastructure upgrades and redesign.

This tutorial was last tested on June 18, 2026, using ArcGIS Pro 3.7. If you're using a different version of ArcGIS Pro, you may encounter different functionality and results.

Requirements

• ArcGIS Pro (see options for software access)
• ArcGIS Spatial Analyst extension
• Arc Hydro for ArcGIS Pro

Map flood extents for different depths

First, you'll create flood polygons for different flood depths.

Download the data

You'll download the data needed for the tutorial and add it to a project in ArcGIS Pro.

1. Download the ExploreFlooding.zip file.
2. Locate the downloaded ExploreFlooding.zip file on your computer and extract it to a location on your C: drive.

   Note:

   If you extract the data to a location other than your C: drive, the tools you run in this workflow may not work.

3. Start ArcGIS Pro. If prompted, sign in using your licensed ArcGIS organizational account.

   Note:

   If you don't have access to ArcGIS Pro or an ArcGIS organizational account, see options for software access.

4. Under New Project, click Map.

   

5. In the New Project window, for Name, type BearCreekFlooding. Click OK.
6. In the Catalog pane, right-click Folders and choose Add Folder Connection.

   

7. In the Add Folder Connection window, browse to the ExploreFlooding folder that you extracted. Click the folder and click OK.
8. In the Catalog pane, expand Folders and expand ExploreFlooding.

   

   The folder contains BearCreek.gdb, a file geodatabase with vector layers and tables that you'll use.

9. Expand BearCreek.gdb and expand the Layers feature dataset. Right-click the Layers feature dataset and choose Add To Current Map.

   

   The layers are added to the project. Next, you'll add the terrain. The Layers folder contains a digital elevation model that you'll use.

10. Expand the Layers folder. Right-click DEM.tif and choose Add To Current Map.
11. In the Contents pane, right-click DEM.tif and choose Zoom To Layer.

    

    The map zooms to the project area.

    

    Note:

    The symbology of the vector layers you added is random and may differ from the example image. The symbology will not affect the workflow.

    The project contains the following layers:

    • The Culvert layer is point feature layer with a single feature that represents a culvert (an underground channel, or tunnel) through which Bear Creek flows at a place where a road crosses the stream.
    • The Road3D layer contains the road network for the area.
    • The NWM_Flowline layer contains the National Water Model (NWM) stream features in the area.
    • The NWM_Catchment layer contains the areas draining into the NWM_Flowline features in the area.
    • The Bear_Creek_Watershed layer was generated using the Watershed tool.
    • The DEM.tif layer is a digital elevation model of the watershed with a 1-meter resolution, from the USGS 3D Elevation program.

    The organization of the data for this project reflects the following standards for Arc Hydro data:

    • All vector data is stored in a feature dataset (typically named Layers).
    • All tables are stored in the same geodatabase as the vector data.
    • All raster data is stored in a subdirectory (typically named Layers) residing in the project directory where the project geodatabase resides.
    • Raster data is saved in .tif format, unless you are processing very large rasters, for which .crf format might be more efficient.
    • Spatial reference, including horizontal and vertical units, for raster and vector data are the same.

    For additional information about Arc Hydro project best practices, review the Arc Hydro - Project Development Best Practices document from the Arc Hydro help site.

Derive flood polygons

Next, you'll use the Dendritic Batch Process Arc Hydro tool to create flood polygons. This tool applies the Height Above Nearest Drainage (HAND) methodology to generate flood extents for different water depths and creates a rating curve showing the relationship between depth and discharge rate for each stream reach.

1. On the ribbon, click Analysis. In the Geoprocessing group, click Tools.

   

2. In the Geoprocessing pane, in the search box, type dendritic batch process. In the list of search results, click the Dendritic Batch Process tool.

   

   Note:

   If the Dendritic Batch Process tool does not appear in the search results, you may not have Arc Hydro installed. Save the current project and close ArcGIS Pro. Go to the Arc Hydro site and download and install the version of Arc Hydro that matches your version of ArcGIS Pro. After the software is installed, start ArcGIS Pro, open the project, and continue the analysis.

3. For Input DEM Raster, choose DEM.tif.
4. For Input Forecast Stream Feature Layer, choose NWM_Flowline. For Input Forecast Catchment Feature Layer, choose NWM_Catchment.
5. For Input Forecast ID Field, choose ID.

   

   You'll accept the default values for Breach Area Depth Limit, Flood Slice Depth, Number of Flood Slices, Use Healed HAND, and Input Draft Culvert Feature Layer.

   Breach Area Depth Limit is a threshold value to determine which areas needs breaching. Flood Slice Depth and Number of Flood Slices define how many flood polygons will be created and their depth offset interval. Selection of these parameters is a function of the terrain accuracy, complexity, and purpose. The Input Draft Culvert Feature Layer parameter is an optional parameter and is not used here.

6. For Input Road Feature Layer, choose Road3D.

   

7. Click Run.

   Note:

   If the tool fails, it may be because your data is not located on your computer's C: drive.

   The tool runs. It takes about 15 minutes to process the Bear Creek dataset with these parameters.

   Tip:

   You can view tool progress by clicking View Details.

   The part of the tool that takes the most time is the flood slicing process. It iterates through the number of slices defined in the Number of Flood Slices parameter, so more slices will result in longer processing times.

Explore the flood polygons

The Dendritic Batch Process tool creates a file geodatabase with multiple outputs that are important for watershed and floodplain analysis. For flood analyses, it creates the FP.gdb file geodatabase in a folder named FloodDataStack within the standard Arc Hydro analysis folder structure. Next, you'll review the outputs created by the tool.

1. At the bottom of the Geoprocessing pane, click the Catalog tab.

   

2. In the Catalog pane, right-click the ExploreFlooding folder and choose Refresh.

   

3. Expand the ExploreFlooding folder, the FloodDataStack folder, and the FP.gdb file geodatabase.

   

   The FPRiver feature class contains the flood modeling streams derived from NWM_Flowline and the DEM. Each stream segment has a unique ID, stored in the HydroID attribute.

   The FPZoneRiver feature class contains flood slices derived by the tool. These are extent polygons for each flood modeling stream segment (indexed by the StreamID attribute as HydroID of the modeling stream) and each depth of flooding (indexed by the HIndex attribute).

4. Add the FPRiver and FPZoneRiver layers to the map.

   The FPRiver lines do not exactly match the lower-resolution NWM_Flowline features because they are derived from the high-resolution, 1-meter DEM.

5. On the map, zoom in to the eastern section of the watershed, at the area where N Quinlan Park Road crosses Bear Creek.

   

6. In the Contents pane, right-click the color patch for the FPZoneRiver layer and choose Apatite Blue.

   

   Tip:

   To see the name of a color in the color picker, point to it.

   In this area, the road, shown as two lines running in a north-south direction, crosses over the stream. At some flood heights, the road is flooded.

   

   Next, you'll explore the level at which the road floods.

7. In the Contents pane, right-click the FPZoneRiver layer and choose Properties.

   

8. In the Layer Properties window, click the Definition Query tab. Click New definition query.

   

9. Create the query Where HIndex is equal to 2.

   

10. Under the query, click Apply.

    

11. In the Layer Properties window, click Apply.

    

    Don't close the window, because you'll be changing the query expression to visualize different values.

    This query expression filters the FPZoneRiver layer to show only the polygons that represent the flood level HIndex is equal to 2. On the map, the extent of the flood polygon has changed.

    

    HIndex of 2 is the water level where the stream is flooded to a level of 1 meter. HIndex of 1 is the index value for the bottom of the stream. When the stream is running at 1-meter depth in this area, the water passes under the road through a culvert, and the road is not flooded.

12. For the query, click Edit.

    

13. Change the HIndex value to 5 and click Apply. At the bottom of the Layer Properties pane, click Apply.

    

    At the flooding level where the stream is running 4 meters deep (HIndex = 5), the flooded area is larger, but the water still passes through the culvert, and the road is not flooded.

    

14. Edit the query to show polygons where HIndex is 8.

    When the stream is running at 7 meters deep, the road begins to be flooded and becomes dangerous for pedestrians and vehicle traffic.

    

15. Optionally, close the Layer Properties window. Pan the map to the west side of the study area to see the single feature in the Culvert layer.

    

    This location is flooded when HIndex is 8. It is also flooded at the much shallower HIndex value of 3.

    

16. On the Quick Access Toolbar, click the Save Project button.

    

You've generated flood inundation polygons using the HAND methodology and explored flood extents for different water depths. You now have a flood depth library that shows how water would spread at various elevations above the stream. Next, you'll link these water depths to flow frequency values from the National Water Model to create flood probability maps for specific return periods.

Map flood frequency

Previously, you created flood inundation polygons for different water depths. Next, you'll obtain National Water Model (NWM) data that contains predictions of how frequently flows of different volumes will occur on each stream segment. These predictions are expressed as 2-year, 5-year, 10-year, 25-year, and 50-year events.

When you link flow frequency data from the National Water Model retrospective dataset to your flood inundation polygons, you can calculate flood extent and depth of flooding for each downloaded discharge value.

Add tables and raster layers

Arc Hydro has a tool to retrieve the data from the National Water Model and assign the flow frequency values to your stream network. You'll prepare to use the tool by adding certain tables and raster layers to the map.

1. In the Catalog pane, in the FP.gdb geodatabase, right-click the pf_ModelStream table and choose Add To Current Map.

   The table is added to the Standalone Tables section of the Contents pane.

   

   This table is a rating curve table. Rating curves express the relationship between flood stage and stream flow volume. This relationship varies for each stream segment and depends on multiple factors, such as the shape, size, slope, and roughness of the channel. The USGS has additional information about rating curves on theUSGS Stage-Discharge Relation Example page.

   The pf_ModelStream table links flood stage (called H) to the flow (called Q) for each modeling stream (indexed by the RiverID attribute) and each depth of flooding (indexed by the HIndex attribute).

   The rating curve allows calculation of H from a known Q or Q from known H. In this case, you'll be retrieve Q values from the NWM and using the rating curve to calculate H. From H, you can calculate depth and extent of flooding for each modeling stream segment.

2. In the Catalog pane, expand the Layers folder.
3. Add the cat.tif and handhealed.tif rasters to the map. When the Build pyramids window appears, click Yes.

   

   Both of these rasters were created by the Dendritic Batch Process tool when you ran it earlier.

   The cat.tif raster is a catchment raster. It identifies the area draining into each modeling stream segment. The value of each cell in the raster matches the HydroID of the stream segment that the cell drains into.

   The handhealed.tif raster is a HAND raster. The value of each cell represents the height above nearest drainage (HAND), that is, the difference in elevation at that cell from the elevation of the stream cell that the cell flows to.

   Now that these tables and rasters have been added to the map, you're ready to run the tool.

Retrieve the flow frequency data

The Map NWM Retrospect FF Qs Arc Hydro tool queries the National Water Model (NWM) retrospective dataset (Retrospect). It gets flow frequency (FF) data for each flow level (Q) for each modeling stream and maps it onto your data.

1. At the bottom of the Catalog pane, click the Geoprocessing tab. In the Geoprocessing pane, click the Back button.

   

2. In the search box, type map nwm retrospect ff q. In the list of search results, click the Map NWM Retrospect FF Qs tool.

   

3. For Input Streams, choose FPRiver.
4. For COMID Field, choose ID.

   

   This field contains the NWM forecast unique identifier value.

5. Accept the default value for Input Rating Curve Table.
6. For Input HAND Raster, choose handhealed.tif.

   

   You'll accept the default value for Input Catchment Raster.

7. Click Run.

   This tool performs many steps and takes about two minutes to run for the Bear Creek data with these parameters.

8. Click View Details to see the progress messages.

   

   As the tool runs, messages are added to the Messages tab.

   

9. Close the results window.

   When the process is complete, a new feature class, FPRiver_fpzone, is added to the FP.gdb geodatabase and the map.

10. If necessary, in the Catalog pane, expand the ExploreFlooding and FloodDataStack folders and the FP.gdb geodatabase. Add the FPRiver_fpzone feature class to the map.

    

    This layer has a smaller set of polygons than the FPZoneRiver layer you examined earlier.

    The polygons show different flood extents for different flooding return periods. The return periods do not represent a schedule, but a statistical estimate of the average time between flooding events of each magnitude. Short return periods, such as 2-year or 5-year, have a higher probability of occurrence in any given year (1 in 2, or 1 in 5, respectively). Longer return periods, such as 25-year or 50-year, have a lower probability of occurrence in any given year (1 in 25, or 1 in 50).

11. In the Contents pane, right-click the FPRiver_fpzone layer and choose Attribute Table.

    

    Two fields, StreamID and H_Field, contain information relating each of the polygons to a particular stream segment and flood stage return period.

    Several H_Field values are rf_2_h.

    

    This value is the code for the 2-year return period. The code is a compact representation of "return frequency 2 year flood stage". Four polygons, relating to the StreamID of 192, 193, 194, and 195, have this H_Field value.

    Below these rows of the table, the set of StreamID values repeats for the 5-year return period, rf_5_h, and so on, for the other return periods.

12. Close the attribute table.

Map the flood frequency data

Next, you'll make a flood frequency map

1. Right-click the FPRiver_fpzone layer and choose Properties.
2. In the Layer Properties window, on the Definition Query tab, click New definition query.
3. Build the query Where H_Field is equal to rf_50_h.

   

4. Click Apply. At the bottom of the Layer Properties pane, click OK.

   Clicking OK closes the Layer Properties window.

   The layer is now filtered to show the flood zone affected by a 50-year return period flood.

5. In the Contents pane, click the FPRiver_fpzone layer. Click it again to edit the layer name.
6. Type 50-year flood extent and press Enter.

   

7. Right-click the symbol for the 50-year flood extent layer and choose Poinsettia Red.

   

8. In the Contents pane, click the 50-year flood extent layer to select it. Press Ctrl+C.

   The layer is copied.

9. Press the Ctrl+V key combination.

   A copy of the layer is pasted into the Contents pane. You'll edit the copy to show the 25-year flood extent and change the name and symbol color for the layer.

10. Change the name of the copy to 25-year flood extent.
11. Drag the 25-year flood extent layer to just above the 50-year flood extent layer.

    

    The copy of the layer will draw on top of the 50-year layer. This way, the larger extent of the less frequent floods will show around the smaller extent of the more frequent floods.

12. Open the Layer Properties window for the 25-year flood extent layer. On the Definition Query tab, click Edit.

    

13. Change rf_50_h to rf_25_h.

    

14. Click Apply and click OK.
15. Change the color of the 25-year flood extent polygons to Medium Coral Light.

    

16. Optionally, add copies of the layer for the 10-, 5-, and 2-year return frequencies.

    Note:

    Remember to update the name, the definition query, and the symbology. Place more frequent return intervals above less frequent ones, in sequence. For the symbology, make the 10-year flood extent Electron Gold, the 5-year extent Solar Yellow, and the 2-year extent Quetzal Green.

    

17. In the Contents pane, turn off all of the layers except the flood extent layers and the basemap.
18. Navigate to the eastern side of the project area and examine the area where N Quinlan Park Road crosses the stream.

    

    Even for the 50-year return period, the road crossing is not significantly impacted by the flooding.

19. Navigate to the western side of the project area to the location where Fritz Hughes Park Road crosses the stream.

    

    The 25-year and 50-year return interval floods flow over the road at this location.

20. Save the project.

You've retrieved flow frequency discharge values from the National Water Model and created flood maps for various return periods. You can now visualize and analyze flood extents for floods of different magnitudes.

Model road flooding at stream crossings

Previously, you linked flow frequency values to flood extents. Next, you'll analyze a culvert to identify at what flow level the water may flow over the road during flood events.

Examine the culvert

You'll examine the culvert and an associated data table.

1. In the Contents pane, check the Culvert layer to turn it on.

   

2. Uncheck the flood extent layers to turn them off.
3. In the Catalog pane, in the ExploreFlooding folder, expand BearCreek.gdb. Right-click the Culvert_rtc table and choose Add To Current Map.

   

   The Culvert_rtc table is added to the Standalone Tables section of the Contents pane.

   

   This table contains hydraulic analysis results for the culvert that were generated using the Federal Highway Administration (FHA) HY-8 modeling software package. This is a complex hydraulic model that considers multiple types of information about the specific site. It includes physical characteristics of the culvert, such as the number of barrels, the shape of the barrels, and their dimensions. It also considers the road width and shape, and the upstream and downstream stream conditions, including the slope and rating curve for the stream.

   Running this analysis is beyond the scope of this tutorial, but you'll use the results.

4. Right-click the Culvert_rtc table and choose Open.

   This table contains multiple rows of data for a single location, with the CrossingName always being Crossing_150. This crossing is the single point in the Culvert layer on the map, at the location where Fritz Hughes Park Road crosses the stream.

   

   The Culvert_rtc table contains the following other attributes:

   • FeatureID is the HydroID of the culvert feature.
   • H is the depth of water, also called the flood stage.
   • Q is the total flow, also called discharge, that the culvert can convey for a specific depth (H).
   • Q_Barrel is the part of the total flow that flows through culvert barrels, the tunnels or pipes that convey water under the road.
   • Q_Roadway is the part of the total flow that flows over the road, also called overtopping.

   The Q_Roadway value is 0 for H values equal to 1.61 or less. As H values increase, the flow exceeds the capacity of the culvert and Q_Roadway values increase. This represents increasing volumes of water flowing over the road at higher flood stages.

5. Close the Culvert_rtc table.

Add the overtopping flow values

You'll run an Arc Hydro tool to add the hydraulic analysis data to the culvert feature.

1. In the Geoprocessing pane, click the Back button.
2. Search for identify culvert overtopping flow. In the list of search results, click the Identify Culvert Overtopping Flow tool.

   

3. For Input Culvert Layer, choose Culvert.
4. For Input Culvert Rating Curve Table, choose Culvert_rtc.

   

5. Click Run.

   The tool runs and adds the Overtopping_Q attribute to the Culvert feature class.

Compare overtopping discharges

You'll use the detailed culvert analysis data to inform the road overtopping analysis you did earlier, where specific culvert properties were not considered.

1. Open the attribute table for the Culvert layer.
2. Scroll to the end of the table.

   

   • Overtopping_Q is the overtopping discharge at the culvert location.
   • TopofRoadElevation is the elevation of the top of the road at the culvert location.
   • ChannelInvertElevation is the channel elevation at the culvert location.

   The difference between the values of TopofRoadElevation and ChannelInvertElevation is 1.82 meters. This represents the height of the stream impediment that the road represents. Think of it as the height of the dam across the stream caused by the road.

   This information is based on a detailed culvert survey and the hydraulic analysis of the culvert structure. When this information is not available, the flow analysis uses the DEM, which can be less accurate.

   Earlier, you created the FPZoneRiver layer. You saw that at HIndex = 3, where flood water is 2 meters above the channel bottom, the road at the culvert location was flooded.

3. Close the attribute table.
4. In the Contents pane, check the FPRiver layer to turn it on.
5. On the map, click the FPRiver feature near the culvert feature.

   

   The pop-up for the feature shows that its HydroID value is 194.

6. Close the pop-up.
7. In the Contents pane, in the Standalone Tables section, right-click the pf_ModelStream table and choose Open.
8. Scroll down to the row where the RiverID value is 194 and the HIndex value is 3.

   

   The Q value for this row is 49.57. This is the discharge that was previously estimated to flood the road. This value is significantly higher than the 3.91 value obtained through detailed hydraulic culvert analysis. That means that a much lower flow rate actually floods the road than was estimated without the culvert analysis.

   This shows the importance of considering the impact of the road and culvert infrastructure when modeling the local flooding.

9. Close the table. Save the project.

In this tutorial, you created comprehensive flood inundation maps for the Bear Creek Subwatershed by combining HAND terrain analysis with National Water Model flow frequency data. You generated flood extent polygons for different water depths, linked them to flood probability scenarios, and performed a culvert overtopping analysis. These skills provide a foundation for flood risk assessment and emergency response planning.

You can find more tutorials in the tutorial gallery.

Acknowledgements

• The DEM raster was extracted from the 3DEP rasters from the United States Geological Survey (USGS) National Map.
• The National Water Model data (NWM_Catchment, NWM_Flowline, and Bear_Creek_Watershed) was created by the National Oceanic and Atmospheric Administration (NOAA) and hosted by Oak Ridge National Laboratory (ORNL).
• The Culvert and Road3D datasets are used with the permission of Professor David R. Maidment of the University of Texas at Austin.
• World Topographic Map sources: Esri, TomTom, Garmin, FAO, NOAA, USGS, OpenStreetMap contributors, and the GIS User Community