Explore the data and learn about hydro-conditioning
In this first module of the lesson, you will set up your ArcGIS Pro project, explore the data to learn more about the municipality of Hvidovre and its environment, and learn about the concept of hydro-conditioning.
Get set up
First, you need to download the project containing all the data needed in this lesson and open it in ArcGIS Pro. Then you'll review the elements it contains.
- Download the Avedoere_Inundation.zip file and locate the file on your computer.
Note:
Depending on your web browser, you may have been prompted to choose the file's location before you began the download. Most browsers download to your computer's Downloads folder by default.
- Right-click the Avedoere_Inundation.zip file and extract it to a location on your computer, such as your Documents folder.
Next, you'll open the project in ArcGIS Pro.
- Start ArcGIS Pro. If prompted, sign in using your licensed ArcGIS account.
Note:
If you don't have ArcGIS Pro or an ArcGIS account, see options for software access.
This lesson uses tools that will only work with ArcGIS Pro 2.8 or later, as the models presented are not compatible with older versions.
- In ArcGIS Pro, under Open, click Open another project.
- In the Open Project window, browse to the project folder you extracted, click Inundation.aprx to select it, and click OK.
The Inundation project opens. You will review the elements in the project.
- On the ribbon, on the View tab, in the Windows group, click Catalog Pane.
- In the Catalog pane, click the side arrow to expand Databases and Inputs.gdb.
The Inputs.gdb database contains vector and raster data that you'll use during the analysis. The spatial reference for all data is set to UTM Zone32N ETRS89, which is the preferred map projection and horizontal datum for the eastern parts of Denmark. For the vertical datum, all data is set to DVR90 (Danish Vertical Reference 1990). Much of this data has already been added to the Contents pane as layers.
- Collapse Inputs.gdb.
The Databases section also contains two empty geodatabases, OutputsBeforeDike.gdb and OutputsAfterDike.gdb, where you will store the output layers generated during the lesson.
- Collapse Databases. Expand Toolboxes and InundationModels.tbx.
This toolbox contains the models you'll run to perform the analysis.
In this section, you downloaded the project, opened it in ArcGIS Pro, and reviewed the elements it contains.
Explore the data
You'll now explore the data to learn more about the municipality of Hvidovre and its environment.
- Review the current map display.
The municipality of Hvidovre is on the coast of the Baltic Sea, about 8 kilometers southwest of Copenhagen. The southern part of the municipality, outlined in red, constitutes the area of study. The imagery basemap gives you a glimpse of the municipality's layout. In particular, you can see that its eastern and southern sides are in direct contact with the sea.
The municipality government of Hvidovre has deployed great effort to protect the area against storm surges over the years, building a dike in 1964-66 (represented in bright pink) and a second one in 1989 (in cyan blue). It is also interesting to note that a major highway (in dotted red) connects Copenhagen to the island of Amager, where Copenhagen's international airport is located.
A dike is defined as a long wall or embankment built to prevent inundation from the sea. The following picture shows the dike raised in 1964-66 in Hvidovre. It has a height, or crown level, of 3.3 meters.
As you will discover through the analysis, two areas are particularly vulnerable to storm surges. The first one is a residential area on the eastern coast (1) that is situated along a natural salt meadow. The second is an industrial area on the southern coast (2).
The following picture is a view showing the sea, the salt meadow to its left, and the residential area farther behind. The photo is taken a little south of location 1, from the dike that protects the industrial area.
You will now look at the elevation data.
- In the Contents pane, check the check box next to the DHyMSea layer to turn it on.
This raster layer provides elevation information for the area.
Note:
The DHyMSea raster layer is the hydro-conditioned version of a digital terrain model (DTM) that represents Denmark in 2020. Each raster cell gives the elevation above sea level in meters. The raster has a spatial resolution of 0.4 meters, which means that each cell represents a square on the ground of 0.4 by 0.4 meters. Hydro-conditioning is a preparation process that you will learn about in more detail later in the lesson.
- On the map, observe how the elevation varies across the study area. Use the layer's legend in the Contents pane for reference.
The elevation varies from about -4.7 to 20.9 meters. The industrial area on the south side has a very low elevation (dark green), sometimes even lower than the sea level. The residential area on the eastern coast is also rather low (bright green). Other areas of the municipality on the north and northwest sides are clearly located at higher elevations (in yellow, orange, and brown).
You will use the Explore tool to identify specific elevation values.
- On the ribbon, on the Map tab, in the Navigate group, click the Explore downward arrow and choose Selected in Contents.
This setting ensures that the information in the pop-up windows is specific to the layer selected in the Contents pane.
- In the Contents pane, click the DHyMSea layer to select it.
- On the map, click a point covered by the sea.
A pop-up appears, showing that the elevation is 0 (meters).
This elevation value was expected since the DHyMSea layer gives the elevation relative to the sea level.
- Click several points in the industrial area, the residential area on the coast, and the higher elevation neighborhoods and observe how the elevation displayed in the pop-up varies.
The areas with the lowest elevations are likely to be most vulnerable to storm surge inundation. Finally, you will get a sense of the buildings in the study area.
- In the Contents pane, turn on the Buildings layer.
The Buildings layer appears on the map, symbolized in black.
This is a polygon feature class that represents the 2D footprints of residential and industrial buildings in 2020. Note that low elevation areas contain many buildings, which could all be affected in a storm surge inundation.
- In the Contents pane, turn off the Buildings and DHyMSea layers.
In this section, you explored the data to better understand the situation of the municipality of Hvidovre, especially in terms of elevation and proximity to the sea.
Learn about hydro-conditioning
In this workflow, you will model how water would flow throughout the study area at different storm surge levels based on elevation information. To perform such hydrologic analyses, it is important to use a DTM that has first been hydro-conditioned. Hydro-conditioning is the process of making small changes to an elevation raster to ensure that water can flow through it in a realistic manner. This process can be quite complex, as explained in the presentation Creating a Hydrologically Conditioned DEM. However, for now, you'll only learn about a few basics that are essentials for the upcoming analysis.
When the lidar technology captures elevation information, it might not capture certain water flow paths, such as a river passing under a bridge.
If the resulting DTM represents the bridge as a high elevation, it will create an obstacle that will prevent the water from flowing past the bridge. In a case such as this, the hydro-conditioning process must modify the DTM, setting some of the cells that represent the bridge to a lower water-level elevation that will allow the water to flow through. Such a change is called a hydro-adaptation.
You will look at such an example in the Hvidovre elevation raster using a bookmark that was set up to help find the location.
- On the ribbon, on the Map tab, in the Navigate group, click Bookmarks and choose Bike road underpass.
The map zooms in to a location where a bike underpass passes under the freeway from north to south, just along the 1964-66 dike. During a storm surge, the water could flow through the underpass, whose elevation is significantly lower than the freeway. This needed to be represented on the DTM with a hydro-adaptation.
- In the Contents pane, turn on the DHyMSea layer.
The DHyMSea layer displays a thin line of lower elevation (in dark green) that connects the two ends of the underpass: that's the hydro-adaptation that will allow the water to flow through at analysis time.
You will take a look at all the hydro-adaptations implemented in this study area.
- In the Contents pane, right-click DHyMSea and choose Zoom To Layer to go back to the full extent.
- In the Contents pane, turn off the DHyMSea layer and turn on the HydroAdaptations layer.
Line features appear in light pink, marking every hydro-adaptation location.
They might be at sites where the following situations exist:
- A road passes under another road.
- A water stream or drainage channel passes under a road or pedestrian bridge.
- A lower-elevation railroad or road tunnel could allow the water to flow through.
Note:
Once the desired hydro-adaptations were defined as line features, they were then burnt onto the DHyMSea layer, a process explained later in the lesson.
- In the Contents pane, turn off the HydroAdaptations layer.
Another interesting case to consider is the presence of high-water flaps at many locations along the dikes. Several drainage channels traverse the dikes to evacuate the rainwater that falls on land and allow it to flow to the sea. The channels are supplemented with several pumping stations that keep the groundwater levels within the diked areas suitably low. In a sea level surge, high-water flaps enable the closure of the drainage channels to prevent seawater from entering the land. You'll look at an example.
- On the ribbon, on the Map tab, click Bookmarks and choose High water flaps.
As shown on the following example image, a drainage canal (1) flows into the sea (2) under the 1989 dike (marked in bright pink) and through high-water flaps (3).
The following picture shows a close-up of the two high-water flaps viewed from above.
Where such flaps are present, no hydro-conditioning should be added to the DTM, because you want to represent these locations as high-elevation obstacles that will not allow the seawater to flow in.
- In the Contents pane, turn on the DHyMSea layer.
On the map, you can see that no hydro-conditioning was applied at that specific location and the dike elevation remains intact. (The hydro-conditioning line at the lower right of the image represents a bike underpass and is unrelated to the drainage channel.)
Note:
For some other hydrologic analyses, it might make sense to represent these flaps as open, and therefore apply hydro-conditioning to those locations. This is the case in the lesson Model bluespots to map flood risk, which models the behavior of stormwater, and observes whether it can properly run off from the land toward the sea. This illustrates how hydro-conditioning may vary based on the target analysis.
In this first module of the lesson, you set up the ArcGIS Pro project, explored the data, and learned about hydro-conditioning. Next, you will start modeling the potential impact of storm surges for the municipality of Hvidovre.
Model storm surge inundations
You are now ready to start modeling the potential impact of storm surges for the municipality of Hvidovre. You will perform a preliminary investigation of the floodable areas. Then, you'll use the Create Inundation model to realistically map the areas that would be inundated for 16 different storm surge levels. Finally, you will generate layers that summarize that information.
Start investigating floodable areas
If the sea around Hvidovre experiences a storm surge of 1.8 meters, a first approach will assume that the water will reach any areas that are up to 1.8 meters above sea level. Therefore, a first quick approach to evaluate which areas might become flooded in that situation is to select all the cells on the elevation raster with a value of 1.8 meters or lower. You will now do that with the raster function Less Than Equal.
- In the Contents pane, if necessary, click the DHyMSea layer to select it.
- On the ribbon, on the Imagery tab, in the Analysis group, click the Raster Functions button.
- In the Raster Functions pane, in the search bar, type Less Than Equal. Under Math: Logical, click Less Than Equal to open the function.
- Choose the following Less Than Equal parameter values:
- For Raster, choose DHyMSea.
- For Raster2, type 1.8.
Less Than Equal will evaluate each cell from the DHyMSea raster, compare it to the value 1.8 and produce a new output raster. If a cell has a value of 1.8 or lower, the function will return a value of 1. Otherwise, it will return a value of 0.
- Click Create new layer.
The new layer, Less Than Equal_DHyMSea, appears.
Note:
Less Than Equal is a raster function that generates a new raster layer dynamically. This is an efficient approach, but the new layer exists only in the computer's memory. If you remove the layer from the project, it will be gone and you will need to re-create it.
You will change the layer's symbology.
- In the Contents pane, right click Less Than Equal_DHyMSea and choose Symbology.
The Symbology pane appears.
- In the Symbology pane, under Primary symbology, choose Unique Values.
- When the warning appears, asking whether you want to compute unique values, click Yes.
In the Symbology pane, the table of unique values appears. You will click each unique value symbol to change its color.
- Click the color symbol for 0 and choose No color.
- Click the color symbol for 1, and pick a blue color such as Cretan Blue in the color palette.
The layer updates. The area of potential inundation now appears in blue.
- Close the Symbology pane.
- Review the new layer and observe which areas appear inundated.
Most of the study area appears inundated, including the residential area on the eastern coast and the entire industrial area on the southern coast.
- On the quick access toolbar, click Save to save the project.
In this section, you investigated the areas that might become inundated in a storm surge, by selecting all the cells on the elevation raster that have a value of 1.8 meters or lower. While this is an interesting first result, in the next section, you will understand why this approach is too simplistic to produce a realistic inundation map, and apply a more sophisticated approach.
Model storm surges more realistically
In the last section, you identified all the raster cells that are 1.8 meters or lower. This is a good start to model a storm surge, but are all these cells passable? That is, would they actually be inundated during a storm surge? If the water, coming from the sea, encounters a higher-elevation barrier, such as a dike, it will be stopped and will not extend beyond that barrier.
This is illustrated in the following diagram: if the movement of the inundation is from left to right, the area in blue will be flooded, and the area in red will not, because the water will be stopped by the higher elevation barrier.
This means that, once you have identified all the cells lower than the 1.8-meter level, you still need to discover which of these cells are connected so they would be flooded by a storm surge coming from the sea.
To do that, you will use a tool, Create Inundation, that models how water propagates, starting from a source. The source is represented as a feature class that you will turn on now.
- In the Contents pane, turn on the LineAtSea layer.
This line feature (in gray) represents the sea line from where the storm surge will come in the analysis.
In the Create Inundation model, there are two main steps. In the first step, a raster layer is produced for a specific surge level, similarly to what you did using the Less Than Equal tool. Taking the example of a 1.8-meter surge, all cells with values equal to or lower than 1.8 meters could be potentially inundated and are assigned a value of 1. All the cells with values higher than 1.8 meters are treated as nonpassable barriers and assigned a value of NoData.
In the second step, Create Inundation models how the water would propagate starting from the source, taking into account the nonpassable high-elevation barriers. The following diagram illustrates this process:
On the diagram, the cells that could potentially be inundated are marked as 1, and the nonpassable barriers are marked as NoData (ND). Starting from a series of source cells at sea (S), the water propagates from one cell to another, if the cells touch side to side or corner to corner. All the blue cells will progressively get inundated, however, the propagation will be stopped by the nonpassable barrier (ND). As a result, the red cells behind the barrier will not be inundated.
The Create Inundation model, as well as all the other models used in this lesson, were developed with ModelBuilder by the lesson's author at the University of Copenhagen.
Note:
In ArcGIS Pro, ModelBuilder is used to create, edit, and manage geoprocessing models. Geoprocessing models are workflows that string together geoprocessing tools, feeding the output of one tool into another tool as input. ModelBuilder can also be thought of as a visual programming language to create executable workflows.
You will open the Create Inundation model in edit mode to take a look at it.
- In the Catalog pane, expand Toolboxes and the InundationModels toolbox. Right-click Create Inundation and choose Edit.
The model opens.
This model may appear complex, and you don't need to review it in detail in this lesson, but here is some essential information about it:
To find the cells that can potentially be inundated (for instance, under 1.8 meters), the model uses the Set Null tool. To emulate the water propagation process, the model uses the Distance Accumulation tool. That tool is usually used in combination with the Optimal Path as Line tool to discover possible paths through a raster according to specific criteria. In this model, it is used to identify groups of connected cells that can be inundated.
The model uses iteration, so instead of discovering the inundated cells for only one storm surge level (such as 1.8 meters), it can work on an entire series of levels, as specified by the user (1.0, 1.2, 1.4, 1.6, 1.8, and so on). For each storm surge level, it produces several outputs that can serve multiple purposes in further analyses.
As input, the model expects the following:
- A hydro-conditioned DTM (such as the DHyMSea layer)
- A source line feature (such as the LineAtSea layer)
- Specifications for the desired storm surge levels
For every storm surge level, the model's outputs will be the following (the example names are given for the 1.8-meter surge):
- DA180: Distance accumulation raster.
- BD180: Back direction raster.
- Inundated180: Raster showing the inundated area.
- InundatedPoly180: Feature class showing the inundated area as a polygon.
Note:
Optionally, if you want to explore the model in depth, you can zoom in to the model and look at each of its components.
- Close the model. If prompted to save the changes to the model, click No.
You will now run the model with storm surge inundation levels from 1 meter to 4 meters in increments of 0.2 meters. First, you'll set up the default output geodatabase to store all the outputs.
- In the Catalog pane, expand Databases. Right-click OutputsBeforeDike.gdb and choose Make Default.
OutputsBeforeDike.gdb is currently an empty geodatabase. It will be used as the default output for the next several steps of the analysis.
- In the Catalog pane, right-click Create Inundation and click Open to open the model as a tool.
Note:
Alternatively, you can double-click Create Inundation to open it.
- Set the following Create Inundation parameter values:
- For Input DTM raster, verify that DHyMSea is selected.
- For Line At Sea, verify that LineAtSea is selected.
- For Output Workspace, verify that OutputsBeforeDike.gdb is selected.
- Accept the default values for all other parameters.
Note:
The tool has default parameter values that match the names of your current layers and workspace. If you're working with your own data, you would need to specify your own layers and workspace.
Note:
Before you run the model, you should know that this process is computation intensive and might take 30 minutes to 1 hour to execute. If you prefer, you can access output files that have already been generated for you, and stored in a geodatabase that you can download.
- If you choose to run the tool yourself, click Run.
While the process is running, you can click View Details to access status information.
When the process is complete, jump to the next section, Examine the new inundation rasters.
- If you chose to use the ready-made data, download the OutputsBeforeDike_example.gdb.zip file and unzip it into your Avedoere_Inundation project folder.
- In ArcGIS Pro, in the Catalog pane, expand Folders, right-click Avedoere_Inundation and choose Refresh.
The new geodatabase appears. You'll add it to the databases for this project.
- Right-click OutputsBeforeDike_example.gdb and choose Add To Project.
- In the Catalog pane, if necessary, expand Databases, and confirm that OutputsBeforeDike_example.gdb has been added.
In this section, you applied the Create Inundation tool to model more realistically how water would propagate at different levels of storm surge. In the next section, you will examine some of the results.
Examine the new inundation rasters
You now will examine some of the results from the Create Inundation model.
- In the Contents pane, turn off the Less Than Equal_DHyMSea layer and turn on the Buildings layer, to ensure you can see the new results optimally.
- In the Catalog pane, expand Databases and OutputsBeforeDike.gdb (or OutputsBeforeDike_example.gdb, if you chose to not run the model).
- Scroll down to see all the output files generated by the model.
You recognize the names listed earlier (BD, DA, Inundated, and so on) and can see each of them for every storm surge level.
- Right-click Inundated100 and choose Add To Current Map.
The raster appears on the map.
Note:
The color is assigned at random and may vary.
- On the map, observe the area delineated by the Inundated100 raster.
This raster represents the area that would be inundated in a 1-meter storm surge coming from the sea. The dikes acted as barriers and protected the industrial area. In the coastal residential area, which is not protected by a dike, the water comes quite close to the houses, but does not flood any of them.
You will now review the inundation raster for the 1.8-meter surge.
- In the Catalog pane, right-click Inundated180 and choose Add To Current Map.
- In the Contents pane, drag Inundated180 under Inundated100.
This allows you to see how the higher surge is flooding a larger area than the lower surge.
Note:
If you want to choose different colors to display these layers, in the Contents pane, right-click the layer's symbol and choose a new color.
Around the industrial area, the difference is quite small. This is primarily due to the dikes that are successfully stopping the 1.8-meter surge. However, in the coastal residential area, the water has now reached a number of houses. This means that you already identified a weak point of the municipality's protection from storm surges: the fact that no dike protects that coastal residential neighborhood makes it very vulnerable to storm surges.
Note:
The municipality is probably already aware of this issue. The problem is that the residents who have a beautiful view of the salt meadow and the sea don't want to have that view blocked by a high dike. In the future, the community of Hvidovre might need to have a discussion about whether such risks are sustainable, especially in the context of a potential sea level rise due to climate change. However, for now, you will not propose a remediation to this situation.
Next, you will review the inundation raster for the 2-meter surge.
- In the Catalog pane, right-click Inundated200 and choose Add To Current Map.
- In the Contents pane, drag Inundated200 under Inundated180.
The map now displays all three layers.
With this new layer, the change looks dramatic. Most of the industrial area is inundated, as well as large portions of noncoastal residential neighborhoods. This also includes the freeway that goes to the airport, cutting off access to vital infrastructure. This is surprising because the 1964-66 dike was built to withstand storm surges up to 3.3 meters. In a later part of the analysis, you'll investigate to identify the weak point that allowed the water to flow in and enabled the inundation.
- Turn off the Inundated100, Inundated180, Inundated200, and LineAtSea layers.
- Press Ctrl+S to save the project.
In this section, you used the Create Inundation tool to model more realistically how storm surges of various heights would propagate from the sea. An initial review of the results allowed you to identify some serious vulnerabilities to storm surges in the Hvidovre municipality.
Summarize the results
While it was useful to look at some of the inundation rasters one by one, it would be better to summarize the information provided by all 16 inundation rasters into a single raster. You'll use the Cell Statistics geoprocessing tool to generate a raster showing the minimum surge height at which each cell will be inundated.
- On the ribbon, in the Analysis pane, in the Geoprocessing pane, choose Tools.
The Geoprocessing pane appears.
- In the Geoprocessing pane, search for and open the Cell Statistics (Spatial Analyst) tool.
The input to that tool will be all 16 Inundated raster layers.
- On the Cell Statistics Parameters tab, for Input rasters or constant values, click Browse.
- In the Input rasters or constant values window, expand Databases and OutputsBeforeDike.gdb (or OutputsBeforeDike_example.gdb, if you chose not to run the model).
- Pressing the Shift key, click Inundated100 and Inundated400 to select all 16 layers. Click OK.
- Choose the remaining parameter values:
- For Output raster, type MinimumInundationBeforeDike.
- For Overlay statistic, choose Minimum.
- Ensure that Ignore NoData in calculations is checked.
To build the output raster, for each cell, Cell Statistics will look in all 16 rasters, and choose the lowest value it finds.
You only want to see the inundation results on the land, and not on the sea, so you'll set up a mask that limits the Cell Statistics processing to the relevant land surfaces.
- Click the Environments tab. Under Raster Analysis, for Mask, expand the drop-down list and choose LandPolygon.
- Click Run.
The new raster layer appears. You'll change its symbology to make it more meaningful.
- Right-click MinimumInundationBeforeDike and choose Symbology.
- In the Symbology pane, for Color Scheme, expand the drop-down list and check Show names, scroll down, and choose the Yellow-Green-Blue (Continuous) color ramp.
On the map, the layer updates. The imagery basemap is showing underneath, and the Buildings layer is displayed on top.
- Close the Symbology pane.
You will examine some of the layer's values.
- In the Contents pane, click the MinimumInundationBeforeDike layer to select it.
- On the map, click some cells of the MinimumInundationBeforeDike layer to display the pop-up information and identify their minimum inundation level.
For example, in the residential area on the eastern coast, you can observe that several residential buildings are in danger of already getting inundated by a 1.4-meter storm surge. The layer also confirms that the industrial area within the 1964-66 dike gets inundated by a 2.0-meter surge. More generally, you can now tell the exact storm surge level that will inundate any cell in the layer.
In this section, you generated a raster layer that summarizes the 16 inundation rasters produced by the Create Inundation model.
Assess every building's vulnerability level
Another interesting way to use the information produced by the Create Inundation model is to find out the minimum flood level for each building within the floodable area. This will enable a precise vulnerability assessment at the building level. You'll do that with the Zonal Statistics tool.
- In the Geoprocessing pane, click the Back button.
- Search for and open the Zonal Statistics (Spatial Analyst) tool.
- Enter the following Zonal Statistics parameters values:
- For Input raster or feature zone data, choose Buildings.
- For Zone field, choose OBJECTID.
- For Input value raster, choose MinimumInundationBeforeDike.
- For Output raster, type BuildingRiskLevel.
- For Statistics type, choose Minimum.
- Ensure that Ignore NoData in calculations is checked.
As some buildings are on terrain with a slope, they may overlap several inundation levels. You are using the Minimum option to ensure that the lowest inundation level possible is selected for each building.
- Click Run.
The new layer appears. You'll turn off some layers to make it easier to see and change its symbology.
- In the Contents pane, turn off the Buildings and MinimumInundationBeforeDike layers.
- Right-click the BuildingRiskLevel layer and choose Symbology.
- In the Symbology pane, for Color scheme, scroll down and choose the Yellow-Green-Blue (Continuous) color ramp.
The layer updates, showing the same symbolization as the MinimumInundationBeforeDike layer.
The layer shows the minimum level (in centimeters) where each building gets inundated.
- Close the Symbology pane.
- In the Contents pane, select the BuildingRiskLevel layer.
- On the map, click some of the buildings in the residential neighborhood on the eastern coast to see their pop-up information.
Once again, it appears that these buildings are the most vulnerable, as they will be flooded at levels between 1.4 and 1.6 meters. This means that many of these buildings must have been affected during the storm in December 2013, when a surge level of 1.64 meters was reported in Copenhagen's harbor, located only 6 kilometers north of this area.
- Continue to explore the layer to understand the status of buildings in other neighborhoods.
- Press Ctrl+S to save the project.
Note:
Optionally, to further assess the potential damage to buildings or other features at various inundation levels, you could run the Inundation Depths model (also provided in the InundationModels.tbx toolbox). Starting from the Create Inundation results, Inundation Depths will generate new rasters containing the inundation depth for every cell (in meters). This is done by subtracting the DTM's elevation values from the inundation raster values.
As you did with the Cell Statistics tool, it is recommended that you set a land mask such as LandPolygon on the tool's Environments tab, to get the inundation depths only on land.
In this module, you modeled the potential impact of storm surges for the municipality of Hvidovre. You first investigated floodable areas looking only at cell elevation. You then used the more sophisticated Create Inundation model to generate realistic inundation maps for 16 different levels of storm surge. Finally, you generated a summary raster layer as well as a vulnerability assessment at the building level. Next, you will use these inundation models to search for inundation entry points.
Search for inundation entry points and add a dike
In the previous module, you saw that a 2.0-meter storm surge would have dramatic consequences for the Hvidovre municipality. You now would like to understand why that 2.0-meter storm surge would produce such an extensive inundation of the industrial area and other neighborhoods. As the existing dikes were built to stop water up to a level of 3.3 meters, you suspect that the weak point must be at a location where there is no dike. You'll identify that inundation entry point, then propose the construction of a new dike to eliminate it. Finally, you'll run the Create Inundation model again, to verify that the proposed new dike fully solves the problem.
Identify the inundation entry point
You will identify the inundation entry point with the Trace Inundation model. From choosing a point in the middle of the inundated area, the model will determine the optimal path that the water might have taken to get to that point. The entry point should be located somewhere along that path.
This model uses data that was generated by the Create Inundation model, more specifically the distance accumulation and back direction rasters. Since you are investigating the water's behavior for a 2.0-meter storm surge, the input will be DA200 and BD200.
You will first take a look at the Trace Inundation model.
- In the Catalog pane, expand Toolboxes and InundationModels.tbx. Right-click Trace Inundation and choose Edit.
The model opens in Edit mode.
The model relies primarily on the Optimal Path As Line tool, which calculates the optimal path from a source to a destination as a line.
- Close the Trace Inundation model. If prompted to save the changes to the model, click No.
Next, you will run the model.
- In the Catalog pane, double-click Trace Inundation.
The Trace Inundation model opens in tool mode. The tool allows you to interactively choose the point that will be back-traced. First, you'll optimize the map display for this action.
- In the Contents pane, turn off BuildingRiskLevel and turn on MinimumInundationBeforeDike to better see the area of interest.
- Turn on the BacktracePoint layer.
This layer indicates a good location for the Backtrace point, to be used as a guide.
- In the Trace Inundation pane, for Backtrace Point, click Create new features in the current map to use as input and choose Points.
- On the map, hover over the red point and click it.
- On the editing toolbar, click Finish to complete the point.
The point is saved in the new layer, Trace Inundation Backtrace Point (Points). The back direction and distance accumulation rasters are not in your map as layers, so you will get them from the output geodatabase.
- Enter the remaining parameter values for the Trace Inundation tool:
- For Back Direction Raster, click Browse, browse to Databases > OutputsBeforeDike.gdb (or OutputsBeforeDike_example.gdb), select BD200, and click OK.
- For Distance Accumulation Raster, click Browse, browse to Databases > OutputsBeforeDike.gdb (or OutputsBeforeDike_example.gdb), select DA200, and click OK.
- For Output Workspace, verify that OutputsBeforeDike.gdb is selected.
- For Output Inundation Flow Path, type InundationPath200 at the end of the %Output Workspace%\ path.
Note:
Ensure that you do not leave spaces before or after InundationPath200.
- Click Run.
When the process is complete, you will add the new layer to the map.
- In the Catalog pane, under Databases, right-click OutputsBeforeDike.gdb and choose Refresh. Right-click InundationPath200 and choose Add To Current Map.
The new layer appears, containing the path found by the model. You will change its symbology to make it more visible.
- In the Contents pane, click the InundationPath200 symbol.
- In the Symbology pane, under Gallery, choose the 2.5 Point style.
The layer updates.
- Close the Symbology pane.
- In the Contents pane, turn on the LineAtSea layer. On the map, zoom in to better see the path feature.
As shown on the following image, the path feature indicates how the water propagated from the gray LineAtSea source (1) to your point of interest in the industrial zone (4). Two other locations along the way (2 and 3) are of particular interest.
At location 2, the water entered the industrial zone just north of the 1964-66 dike (bright pink). This seems to be the weak point you were looking for. At location 3, it seems that water was able to pass the freeway by going through a low-elevation bike underpass.
The inundation entry point you discovered is located where the 1964-66 dike ends. In that area, the terrain suddenly decreases from 3.3 meters to 1.8 meters.
Note:
Optionally, you could verify this sudden decrease on the DHyMSea layer.
To better understand what might be happening at the entry point, you can look at a corresponding field photo.
The dike with a crown level of 3.3 meters ends abruptly (1) and the terrain slopes down to a level of 1.8 meters where the boulders are located (2). However, the photo reveals that a narrow concrete wall takes over where the dike ends (3). This wall is very thin (about 20 centimeters wide) and it is partially covered by overhanging vegetation. So, it was not captured in the DTM, which has a spatial resolution of 40 centimeters.
Note:
This is a good example of how narrow features might not be captured, even by high-resolution lidar technology, but might be revealed during field inspections. This is also a reminder that the findings of such a GIS workflow should always be verified by a field inspection.
In a real-life setting, the next step might be to ask engineers to assess the thin wall. The wall was probably established to withstand surges above the 1.8-meter level, and engineers might conclude that the wall is still strong enough. Or they might recommend that a stronger wall or a dike extension should be considered in times of climate change.
In the context of this lesson, you will assume that a dike extension must be built, and examine this scenario in the next several sections.
In this section, you used the Trace Inundation model to find the inundation entry point that would produce an extensive inundation in a 2.0-meter storm surge. You then inspected a field photo to better understand the situation on the ground.
Create a dike feature class
Now that you revealed the inundation's entry point, you want to propose the construction of a new dike to better protect the municipality of Hvidovre. First, you will create an empty line feature class where you will later store the new dike feature.
- In the Catalog pane, expand Databases. Right-click Inputs.gdb, and choose New > Feature Class.
- In the Create Feature Class pane, on the Define page, choose the following property values:
- For Name, type NewDike.
- For Feature Class Type, choose Line.
- For Geometric Properties, uncheck Z Values.
Note:
The feature class does not need to be 3D, as the elevation will be stored in an attribute.
- Click Next.
You now need to create a numeric field named Z to store the elevation value of the new dike (expressed in meters).
- On the Fields page, click Click here to add a new field, and enter the following values:
- For Field Name, type Z.
- For Data Type, choose Float.
The Float type stores numeric values with decimals.
- Click Next. On the Spatial Reference page, verify that the Current XY option is set to ETRS 1989 UTM Zone 32N.
This is the coordinate system used in the project.
- Click Finish to complete the creation of the feature class.
The new feature class appears in the Contents pane; it is currently empty.
In this section, you created a line feature class named NewDike.
Digitize the proposed new dike
In this section, you will digitize the proposed new dike as a line feature in the NewDike feature class, and store its elevation as an attribute. First, you will zoom in to the location for the proposed new dike.
- On the ribbon, on the Map tab, click Bookmarks and choose Dike Extension.
- In the Contents pane, turn on the Buildings layer and ensure that the MinimumInundationBeforeDike layer is also turned on.
You should carefully consider where to place the dike and what elevation to assign as its crown level. In this case, it should function as an extension of the 1964-66 dike: it should start from the existing dike, go in the northwest direction, and end where the terrain becomes naturally elevated. In terms of elevation, it should stop 3.3-meter surges, just as the existing dike. The end result will appear as represented with a red dashed line on the following example image:
To decide where to stop the line, you need to perform some readouts on the map.
- In the Contents pane, click the MinimumInundationBeforeDike layer to select it.
- On the map, click several points along the desired line to view the pop-ups and identify where the minimum inundation reaches the value close to 3.3 meters.
This is where you should stop the line, as that area corresponds to grounds that are naturally more elevated.
You want to ensure that the new dike feature touches the old dike; otherwise, it will leave space for the water to flow through. You will turn on Snapping for that purpose.
- On the ribbon, on the Edit tab, click the Snapping down arrow to expand the Snapping pane.
- In the Snapping pane, click the following two buttons to turn them on:
- Turn snapping on or off for interactive tools.
- Endpoint snaps to the nearest start or endpoint of a polyline feature.
When turned on, the two buttons appear blue.
- Click the Snapping down arrow to close the Snapping pane.
You will now start digitizing the proposed new dike.
- On the ribbon, on the Edit tab, in the Features group, click Create.
- In the Create Features pane, click NewDike and ensure that the Line tool is selected.
- On the map, point to the Dike1964 endpoint, until the snapping marker appears.
This will be the beginning of your dike extension.
- On the map, click the snapping marker, then double-click the desired endpoint of your dike extension.
The new dike feature appears.
Note:
For the purpose of this exercise, it is not an issue if the new dike slightly overlaps with a few existing buildings.
- On the floating editing toolbar, click Finish.
- On the Edit tab, in the Selection group, click Clear to deselect the new feature.
Note:
Optionally, you can change the symbolization of the new layer to better see the new dike feature.
- On the ribbon, on the Edit tab, click the Snapping button to turn Snapping off.
You now need to add the elevation of the new dike. It should be 3.3 meters to match the existing dike.
- In the Contents pane, right-click the NewDike layer, and choose Attribute Table.
In the attribute table, the new dike feature is listed as the only row.
- In the row, click the Z field and type 3.3.
You will save the new feature.
- On the ribbon, on the Edit tab, in the Manage Edits group, click Save.
Note:
When working with your own data, you could add several dikes in different locations. You would create all of them in the same NewDike layer. You could assign different elevations to each dike by using the Z attribute in the attribute table. You could also digitize a single dike as a line with several vertices, instead of just making it a straight line.
- In the Contents pane, right-click MinimumInundationBeforeDike and choose Zoom To Layer to go back to the full extent.
- Close the Create Features pane and the NewDike attribute table.
- Press Ctrl+S to save the project.
In this section, you digitized the proposed new dike as a line feature.
Create a map
The rest of the analysis will focus on modeling inundation levels with the proposed new dike. For clarity, you will create a map of where to execute that part of the analysis. You'll make a modified copy of the current map.
- In the Catalog pane, expand Maps. Right-click Coastal Inundation and choose Copy.
- Right-click Maps, and choose Paste.
A new map, Coastal Inundation1, appears.
- Right-click Coastal Inundation1, choose Rename, type After Dike, and press Enter.
The new map's name updates in the Catalog pane.
You will now open the map and remove any layers that are not relevant for the rest of the analysis.
- Right-click After Dike, and choose Open.
The new map opens.
- Ensure that the After Dike map tab is selected.
- In the Contents pane, right-click Trace Inundation BacktracePoint (Points) and choose Remove.
- Similarly, remove the following layers:
- InundationPath200
- Inundated100
- Inundated180
- Inundated200
- MinimumInundationBeforeDike
- Less Than Equal_DHyMSea
- In the Contents pane, turn off BacktracePoint.
Keep other layers such as NewDike and Dike1964 on for context. The new map should look like the following example image:
In this section, you created a map, where you will execute the rest of the analysis.
Burn the proposed new dike onto the elevation raster
Now that you have created a line feature representing the proposed new dike, you want the DHyMSea elevation raster to reflect this new element. This is done by burning the dike onto the raster, changing the corresponding cells to a 3.3-meter elevation value. You will burn the dike onto the DHyMSea layer using the Burn Dike Onto DTM model.
- In the Catalog pane, expand Toolboxes, right-click the Burn Dike Onto DTM model, and click Edit to open it in edit mode and inspect it.
The model expects the following:
- A dike feature layer or feature class input
- A DTM raster input
- A workspace location and a name for the resulting DTM output
First, the Polyline To Raster tool produces a raster, DikeRaster, where the dike feature is represented with cells of Z elevation value. Then the model uses Raster Calculator expressions to replace the relevant DHyMSea cells with the DikeRaster cell values.
- When you are done inspecting the model, close it.
- In the Catalog pane, double-click the Burn Dike Onto DTM model to open it in tool mode.
- Enter the following Burn Dike Onto DTM parameter values:
- For Input DTM, verify that DHyMSea is selected.
- For Dike line feature class or layer, verify that NewDike is selected.
- For Dike Z value field, verify that Z is selected.
- For Output Workspace, click Browse, expand Databases, select Inputs.gdb, and click OK.
- For Output DTM with Dike, and type DHyMSeaWithDike at the end of the %Output Workspace%\ path.
Note:
You want to store the new DTM layer in Inputs.gdb next to DHyMSea.
- Click Run.
When the process is complete, you will add the new DHyMSeaWithDike raster to the map and verify that it reflects the dike extension.
- In the Catalog pane, expand Databases > Inputs.gdb. If necessary, right-click Inputs.gdb and choose Refresh.
The new raster appears listed under Inputs.gdb.
- Right-click DHyMSeaWithDike and choose Add To Current Map.
- If asked whether you would like to build pyramids for the raster, choose Yes.
The new raster appears on the map.
- In the Contents pane, turn the NewDike and Buildings layers off so that you can better see the raster.
- On the ribbon, on the Map tab, click Bookmarks and choose Dike Extension. Zoom in more if necessary.
You can see the thin line of the extension dike.
- In the Contents pane, right-click DHyMSeaWithDike and choose Zoom To Layer to go back to the full extent.
Note:
Optionally, you could symbolize the DHyMSeaWithDike raster similarly to DHyMSea, using the Elevation #6 color ramp.
- Press Ctrl+S to save the project.
In this section, you burnt the proposed new dike extension onto the elevation raster.
Rerun the inundation model
You are now ready to rerun the Create Inundation model to see whether your proposed dike extension can successfully stop inundations for storm surges up to 3.3 meters. You want to store the results for the rest of the analysis in the OutputsAfterDike.gdb geodatabase, so you'll first make it the default.
- In the Catalog pane, expand Databases. Right-click OutputsAfterDike.gdb and choose Make Default.
You will now run the model.
- In the Catalog pane, expand Toolboxes, and double-click Create Inundation to open it.
- Enter the following Create Inundation parameter values:
- For Input DTM raster, choose DHyMSeaWithDike.
- For Line At Sea, verify that LineAtSea is selected.
- For Output Workspace, click Browse, select OutputsAfterDike.gdb, and click OK.
- Accept all the other default values.
Note:
Like the first time, running this process is computation intensive and might take 30 minutes to an hour to execute. If you prefer, you can access output files that have already been generated for you, and stored in a geodatabase that you can download.
- If you chose to run the tool, click Run.
When the process is complete, jump to step 7.
- If you chose to use the ready-made data, download the OutputsAfterDike_example.gdb.zip file and unzip it into your Avedoere_Inundation project folder.
- InArcGIS Pro, in the Catalog pane, expand Folders, right-click Avedoere_Inundation and choose Refresh. Right-click OutputsAfterDike_example.gdb and choose Add To Project.
In the Catalog pane, under Databases, OutputsAfterDike_example.gdb appears.
You will now summarize the results by generating a MinimumInundationAfterDike raster using the Cell Statistics tool. It will be similar to the MinimumInundationBeforeDike raster you produced earlier.
- In the Geoprocessing pane, click the Back button. Look for and open the Cell Statistics tool.
- In the Cell Statistics Parameters pane, for Input rasters or constant values, click Browse.
- In the Input rasters or constant values window, expand Databases and OutputsAfterDike.gdb (or OutputsAfterDike_example.gdb, if you chose to not run the model).
- Pressing the Shift key, click Inundated100 and Inundated400 to select all 16 layers. Click OK.
- In the Cell Statistics Parameters pane, choose the remaining parameter values:
- For Output raster, type MinimumInundationAfterDike.
- For Overlay statistic, choose Minimum.
- Ensure that Ignore NoData in calculations is checked.
- Click the Environments tab. Under Raster Analysis, for Mask, expand the drop-down list and choose LandPolygon.
- Click Run.
The new raster layer appears. You will change its symbology and customize the map's general appearance.
- In the Contents pane, right-click MinimumInundationAfterDike and choose Symbology.
- In the Symbology pane, for Color Scheme, expand the drop-down list and choose the Yellow-Green-Blue (Continuous) color ramp.
- In the Contents pane, turn off the DHyMSeaWithDike layer, and turn on the NewDike and Buildings layers.
The map updates.
- Close the Symbology pane.
You can see immediately that the situation of the industrial area has greatly improved, as it is now displayed with a darker shade of blue. You will look at a few specific minimum inundation values.
- In the Contents pane, ensure that the MinimumInundationAfterDike layer is selected.
- On the map, click some cells of the MinimumInundationAfterDike layer to display the pop-up information.
The area inside the 1964-66 dike, which was previously inundated by a 2.0-meter surge, will now only be inundated at 3.0 meters. This is a great improvement, so your proposed dike seems quite successful in remedying the issue. However, this result is still not as high as expected, since the original dike as well as yours are supposed to be built to withstand a 3.3-meter surge. Is the dike extension not long enough, or are there other weaknesses along the dike from 1964-66 that you have not yet discovered?
- Press Ctrl+S to save the project.
In this section, you modeled inundations on the modified elevation raster to evaluate the efficacy of your proposed dike extension. In the next section, you'll finalize the analysis by investigating the reasons for the 3.0- to 3.3-meter discrepancy you identified.
Trace the water propagation path
To understand what causes the industrial area to get inundated at 3.0 meters instead of 3.3 meters, you will generate the water propagation path on the new data with the Trace Inundation model, focusing on the 3.0-meter surge.
- In the Contents pane, turn on the BacktracePoint for guidance.
- In the Catalog pane, expand Toolboxes, and double-click the Trace Inundation model to open it.
- In the Trace Inundation pane, for Backtrace Point, click Create new features in the current map to use as input and choose Points.
- On the map, hover over the red point and click it. On the editing toolbar, click Finish to complete the point.
The point is recorded in the new layer, Trace Inundation Backtrace Point (Points).
- Enter the remaining parameter values for the Trace Inundation tool:
- For Back Direction Raster, click Browse, browse to Databases > OutputsAfterDike.gdb (or OutputsAfterDike_example.gdb), select BD300, and click OK.
- For Distance Accumulation Raster, click Browse, browse to Databases > OutputsAfterDike.gdb (or OutputsAfterDike_example.gdb), select DA300, and click OK.
- For Output Workspace, click Browse, expand Databases, select OutputsAfterDike.gdb , and click OK.
- For Output Inundation Flow Path, type InundationPath300AfterDike at the end of the %Output Workspace%\ path.
- Click Run.
When the process is complete, you will add the new layer to the map.
- In the Catalog pane, under Databases, right-click OutputsAfterDike.gdb and choose Refresh. Right-click InundationPath300AfterDike and choose Add To Current Map.
The new layer appears, containing the path found by the model. You will change its symbology to make it more visible.
- In the Contents pane, click the InundationPath300AfterDike symbol.
- In the Symbology pane, for Color, choose a bright yellow, such as Solar Yellow.
- For Line width, type 3 pt.
The layer updates.
- Close the Symbology pane.
- In the Contents pane, turn off MinimumInundationAfterDike and Buildings to better see the imagery basemap.
This time, the inundation path seems to indicate that there is an issue along the 1964-66 dike, just south of the freeway where it turns into a bridge toward the island of Amager.
- Zoom in further into that area.
According to the imagery basemap, the inundation seems to enter through a low-lying area, close to the bike lane underpass you investigated earlier. You will investigate this further in the next section.
- Press Ctrl+S to save the project.
In this section, you used the Trace Inundation model one more time to generate the water propagation path on the new inundation data and the case of a 3.0-meter surge.
Generate a profile graph for the 3D water path
To better understand the elevation variations in that area, you will generate a cross-section representation of the elevation values along the InundationPath300AfterDike path.
First, you use the Interpolate Shape tool to create a 3D line feature based on the InundationPath300AfterDike path, and the DHyMSeaWithDike elevation values it encounters along the way.
- In the Geoprocessing pane, click the Back button. Search for and open the Interpolate Shape (Spatial Analyst) tool.
- Enter the following Interpolate Shape parameters:
- For Input Surface, select DHyMSeaWithDike.
- For Input Features , select InundationPath300AfterDike.
- For Output Feature Class , type InterpolatedCrossSection.
- Click Run.
The new layer, InterpolatedCrossSection, appears. Currently, it looks very similar to InundationPath300AfterDike, because it is seen from above. However, this is a 3D line that goes up and down according to the elevation information found in the terrain. A good way to visualize it is to generate a cross section as a Profile Graph.
- In the Contents pane, right-click InterpolatedCrossSection, choose Create Chart > Profile Graph.
The Profile Graph pane and Chart Properties pane appear. First, you will change the graph's color to better see it.
- In the Contents pane, click the InterpolatedCrossSection symbol.
- In the Symbology pane, for Color, choose a dark green color, such as Leaf Green.
- For Line width, type 3 pt.
The color for InterpolatedCrossSection updates on the map and on the chart.
You notice that on the right side of the graph, a high plateau seems to represent the elevated dike. It is followed by a strong dip, which might coincide with the bike underpass. To verify this is the case, you will use the Filter by Extent capability.
- On the Profile Graph ribbon, for Filter, click Extent.
- On the map, zoom in to the area of interest.
The elevated plateau corresponds indeed to the part of the dike that runs along the highway, just before it dips to give way to the bike underpass. Next, you want to read the elevation of that dike area.
- Point to several locations on the elevated plateau area on the graph.
A pop-up displays Elevation values, and the highest value seems to be just under 3.0 meters.
This verifies your finding that a 3-meter surge could flow through this point of entry, and it does seem caused by a slightly lower dike in that area. Maybe some municipality officers should inspect the location and consider how to raise the level by 30 centimeters.
Note:
Alternatively, if the municipality decides to not make a permanent improvement in that area, the disaster management team will know where to reinforce the dike with a mobile barrier during a critical surge level.
- Press Ctrl+S to save the project.
In this section, you generated the cross-section representation of the elevation values along the InundationPath300AfterDike path, and you determined that the entry point was through a segment of the 1964-66 dike that is a bit lower than the expected 3.3 meters.
In this module, you searched for inundation entry points, proposed a new dike, and evaluated its efficacy. You have now completed the workflow proposed in this lesson. In the next module, you'll get some hints on how to run a similar analysis for your own region of interest and with your own data.
Model storm surge inundations in your region of interest
In this last module, you'll review some information to help you apply the workflow described in this lesson to your own region of interest.
Understand the strengths and limitations of the models
First are a few general considerations about the models used in the lesson.
The models are primarily meant to study the impact of coastal inundations induced by storm surges. The ability to incorporate proposed new dikes is useful when considering the implementation of long-lasting shoreline protections. Some of the results can also be used in disaster management operations (as indicated in the study of the cross section): they can indicate where to roll out mobile barriers based on forecasted storm surges.
Note:
The workflow has also proven useful to detect and correct potential errors in a DTM, such as a thin wall whose elevation is not represented.
The inundation raster outputs from the Create Inundation models are provided not only as rasters but also as feature classes showing the inundation extents. These feature classes can provide opportunities for further investigations, such as producing overlay analyses to identify the type of land-use categories affected. The various raster outputs, predicting the depths at which specific infrastructures will be flooded, can be used to estimate damage costs. If generated in advance, they are ready to use quickly in case of a disaster management situation.
Of course, the models used in this lesson have limitations, as they do not account for all the complexity of real-life coastal inundation dynamics. Real storm surges generate tall waves that are higher than the average surge level. As a result, these high waves can splash over the land at lower levels than modeled. Furthermore, along coastlines, the erosion of dunes or other unconsolidated materials may cause inundations at lower levels than predicted. The models used are deterministic, which means that they assume steady state conditions, such as stable elevation (for instance, they do not consider that a water barrier could collapse during a storm surge).
In the context of climate change, the models could also be useful to identify the coastal areas that are in danger of getting permanently inundated due to steady sea level increases. However, in dynamic coastal areas where erosion or accumulation rates are high, the landscape is subject to changes in elevation levels over time. To fully take into account these dynamics, more complex modeling would be needed.
Also, the models cannot be used for time-related predictions: they cannot forecast when a specific location gets inundated during a storm. This would require the use of 2D hydro-dynamic software.
However, the models presented here are excellent to provide a first overview of storm surge consequences, to control the quality of a hydro-conditioned DTM, and for identifying possible remedies to inundations.
Choose the right DTM data and extent
The most important data you need to run this inundation workflow on your own area of interest is a DTM. A high-resolution DTM, for instance derived from lidar, is ideal. This is the case for the data used in the lesson, which is based on a 0.4-meter DTM from 2020. However, if you only have access to a lower-resolution DTM, do not hesitate to still experiment with it. The results will be less precise but can still be worthwhile.
The vertical accuracy for the DTM used in this lesson is very high, close to 0.05 meters. However, a lower vertical accuracy is acceptable too.
It is very important that the extent chosen for your analysis be sufficiently large and it should encompass the complexity of water movements around bends, spits, and other landforms. For instance, a storm surge could inundate an ancient lagoon behind a coastal barrier by flowing through a water entry point that is quite remote from the immediate area of interest. If the extent you choose is too small, it might miss some of these dynamics.
Note:
The size of the extent chosen in this lesson to study Hvidovre's vulnerability is not a good example. It was kept artificially small to limit the size of the data package and keep the computation times low. The analysis was first run on a larger extent. Then, based on the results, a smaller extent that captured all the interesting dynamics was chosen. In a real-life setting, it would be important to choose an area that extends well beyond the municipality of Hvidovre.
Here are a few more recommendations:
- Your data must be in metric units and it must be in a projection suitable for your area. If necessary, use the Project Raster tool to change your data's projection.
- When digitizing the LineAtSea line feature class, ensure that it is located within the DTM's extent.
- It is recommended that you use a land mask, such as LandPolygon, to limit certain results to land areas, for instance, when using the Cell Statistics and Inundation Depths tools.
- However, the land mask must not be used with the CreateInundation tool, which must contain sea-covered cells to work properly.
- It is useful to use a similar geodatabase structure as the one used in this lesson, distinguishing Inputs.gdb, OutputsBeforeDike.gdb, and OutputsAfterDike.gdb.
Hydro-conditioning: burn flow lines onto your DTM
As you learned in the Learn about hydro-conditioning section, it is crucial that you hydro-condition your DTM raster before performing the analysis. One type of hydro-adaptation that is essential for an inundation analysis is to add missing flow paths to your DTM to represent water streams flowing under bridges, road underpasses, railroad tunnels, and so on. Below are some hints on how to do this yourself.
You first need to create an empty line feature class, named HydroAdaptations, for instance. Then you should digitize the desired hydro-adaptations as feature lines and store them in the HydroAdaptations feature class. This process is mostly similar to what you did in the Digitize the proposed new dike section, with the difference that you would not be adding dikes but flow paths. For instance, a flow line added to a bike underpass would look like this, as seen earlier in the lesson:
Next, you need to burn all the hydro-adaptations stored in the HydroAdaptations layer onto the DTM. This can be done with the Burn Flowlines Onto DTM tool in the InundationModels.tbx toolbox.
For each hydro-adaption line feature, the model locates the lines' endpoints, assigns them elevation values extracted from the DTM and averages them. This mean Z value is then automatically assigned to relevant DTM raster cells along the flow line using the Polyline to Raster and Raster Calculator tools. As a result, the water flows are enabled through one-cell-wide hydrologic flow paths.
As an example, you could open the Burn Flowlines Onto DTM model as a tool and enter the following parameters:
- For Input DTM, choose your original DTM.
- For Hydro Adaptations, choose your hydro adaptations layer.
- For Output Workspace, choose your Inputs.gdb geodatabase.
- For Output Flowline Adjusted DTM, type DHyMSea.
As seen earlier, the resulting burnt flow line for the bike underpass will look like this:
Note:
The hydro-conditioning steps described here represent the bare minimum you need to implement for this lesson's workflow to be meaningful. You can learn about more advanced hydro-conditioning approaches in the presentation Creating a Hydrologically Conditioned DEM.
You can also see an example of another hydro-conditioning operation, sink removal, in the Predict floods with unit hydrographs lesson.
When you have hydro-conditioned your DTM, you are ready to apply the inundation screening workflow you learned in this lesson: model storm surge inundations, look for inundation entry points, propose new protective dikes to mitigate the problems, and evaluate their efficacy.
In this lesson, you assessed the vulnerability of the municipality of Hvidovre in Denmark to inundations caused by storm surges. You first determined the potential impact for storm surges of 16 different sea levels varying between 1 and 4 meters. You then identified a weak point that might allow the water to flow through, and proposed the construction of a new protective dike. You verified that the addition of the dike would avoid potential catastrophic inundations in the future and still found a secondary point of concern. Finally, you learned how to replicate this analysis in your own area of interest. This type of inundation screening is very relevant in the context of climate change, sea-level rise, and increased frequency of storm surge occurrences.