Consider creating and using C:\Charlottenlund_Stormwater as a folder location.

The BluespotScreening project opens and displays a map with the imagery basemap and several modeling data layers. You may notice a datum transformation warning, you can ignore this. Before getting started, you may need to verify whether the Spatial Analyst extension is available.

Before proceeding, you will review the current installed Python packages to determine whether the necessary networkx package is available, installed, and configured.

Next, you will verify that the NetworkX Python plug-in is accessible in ArcGIS Pro.

While the cell runs, an asterisk is inside the brackets in the input area so it appears as In [*]. After the cell finishes running, the number 1 replaces the asterisk in the brackets. When the notebook returns 1, it indicates that the package is configured and you may continue.

Next, you will explore the sample data and modeling tools needed to run the bluespot model.

All project data is stored in file geodatabases, and input data is separated from output data in different geodatabases. In addition, the project has several subfolders containing project layer files, scripts, and the model tools.

Note the following:

• The Charlottenlund_Stormwater folder is the project folder (yours may differ).
• The Scripts folder is a location to store project scripts.
• The Inputs.gdb geodatabase contains input sample data.
• The Outputs.gdb geodatabase contains user-generated output data.
• The Results.gdb geodatabase contains results as generated by the modeling.
• The BluespotScreening.tbx file is a toolbox with models and script tools.

The folder contains layer files with symbology ready to import.

The Inputs geodatabase contains the necessary input data to carry out the storm water screening and bluespot modeling.

The project has several bookmarks to assist in locating various landmarks and areas of interest identified in the model results.

Notice that the catchment it is located in the North of Copenhagen, Denmark.

The feature classes in the Inputs geodatabase represent processed data used in the tools and models to predict the flooding consequences for a series of simulated storm water events.

The data is already added to the map as a collection of layers organized in the Contents pane. The layers are organized with line layers displayed above polygon layers and raster layers located below polygon layers.

Next, you will use the Explore tool to pan and zoom and investigate the drainage basin.

The RailroadsWithinDEM and RoadsWithinDEM layers show the locations of all minor and major roads and railroads located within the drainage basin. When comparing your screening results with these layers in the Explore the results lesson, you will be able to identify which parts of the infrastructure are most affected by floods.

The map shows several residential areas containing a mixture of single-family homes and apartment blocks.

During the data preparation phase of the project, a roof was “added” to prevent precipitation from being trapped inside the courtyard. This preparation step will be demonstrated in the Run the analysis with your own data lesson.

Notice that the railroad lines stand out clearly but are only passable through road overpasses or road underpasses built at several locations.

You will explore and investigate some of the critically impacted underpasses in the next lesson, in which you will also consider the flood impacts on residential areas.

The DHyMb layer is a raster layer with cell values representing elevation above sea level in meters for the hydrologically adapted DTM covering the Charlottenlund drainage basin.

While reviewing the terrain model’s sliced elevation values, you may start to get a first impression of where the water moves from the higher elevation boundary of the drainage basin toward the lower elevation dark green region in the central and eastern parts.

The basin has an outlet into the sea east of the Charlottenlund Fort, which was constructed from 1883–1888 with the goal of protecting the entrance into the Copenhagen Seaport.

Notice that buildings are assigned a very high elevation value (30,000) to enforce storm water runoff around them.

Notice that the horizontal spatial reference system for all data used in the map is ETRS 1989 UTM zone 32N, which is a UTM projection with the ETRS89 datum applied. This is the most commonly used spatial reference system for Danish maps with horizontal and vertical units for 3D features in meters.

To ensure that the modeling tool outputs are saved in the correct output workspace, you must make sure that the projects default geodatabase is correctly set.

With the default geodatabase set, results from geoprocessing operations will automatically be directed to this location. Meanwhile, you will notice that some models and script tools include an Output Workspace parameter as well. This is to make sure that outputs from geoprocessing tools referenced as sub-components within other models or scripts are saved in the right location.

At first glance, the model may seem rather complex, but when exploring its various components, you will notice that parts of the model focus on specific aspects of the workflow.

• The left part focuses on determining the surface flow and identifying the bluespots using raster-based tools.
• The central part executes the LocatePourPoints tool, which is a Python script tool to identify the bluespot pour points.
• The right part is mostly about converting the identified bluespots and local watersheds into vector representations, assigning meaningful attribute names to the outputs and joining tables.

Note the sequence of Spatial Analyst tools used to identify the bluespots’ spatial extents by first filling up all landscape depressions detected in the DTM to their pour point levels.

Next, the original elevation values in the Input DEM are subtracted cell by cell from DEMFilled to create a depth raster of all depressions named BSDepths. Afterward, the individual bluespots are assembled by using the Region Group tool.

In addition, notice how a bluespot's volume is calculated by multiplying all cell values in the BSDepths raster by the area of a cell (in the sample dataset, the cell size is .4 * .4 = .16 m²) and then summarizing all values. This is done by using Zonal Statistics and the Volume Calculator tools. The criteria used to ignore small sinks is handled by the Apply Filter tool.

Next, a bluespots pour point is determined by the LocatePourPoints script tool from a search for the cell within a bluespot that has the highest flow accumulation value.

Meanwhile, since the pour point is located along a bluespot’s rim, it isn’t included in the its extent. Therefore, the pour point is moved to the next downstream cell that has a higher flow accumulation value just outside the bluespot using the SnapPourPoint tool. Next, the bluespot is expanded to include the pour point cell, also, before calculating the bluespot’s local watershed using the Watershed tool.

Flow direction and flow accumulation

When flow directions are identified, the flow accumulations describe how the downstream flow merges into minor streams and further downstream into larger river systems. In the illustration below, a DTM (a) is examined for steepest flow directions (b) on a cell-by-cell basis. Next, based on an encoding schema (c), the flow directions become more evident when visualized as arrows (d). The flow accumulation numbers in (e) tell how many upstream cells contribute runoff to a specific cell (e).

On your own, try to verify the flow accumulation number in (e) for the cell with the value 7 by walking upstream in (d).

In the perfect DTM presented above (with elevation values rounded to integers) for a square drainage basin represented by 6x6 cells, all cells contribute to the lower cell with the value 35, and from that one all water leaves in a ‘downward’ direction. In this DTM, there are no local bluespots.

If bluespots were present, the flow directions would end in the lowest lying cell inside it. Therefore, to model a continuous flow for an entire catchment, we fill up all bluespots to their pour point levels using the Fill tool before calculating the flow direction. For a detailed discussion, review the help topic How Fill works.

In the Identify Bluespot features, all landscape depressions are filled based on the Fill tool, which leads to the identification of the desired bluespots.

Finally, the Streams feature class is derived from looking up the steepest downstream paths from all pour points determined by Spatial Analyst’s Optimal Path As Line tool found in the model’s upper right corner.

When using pour points as source cells, flow paths are determined from a "Look up" of the downstream flow directions from cell to cell in the FDFilled raster. The Flow Accumulation raster, FAFilled, is also a required input to this script tool.

The output from the tool is flow paths as line features with downstream directions. When lines from various pour points merge, overlapping lines are turned into single line features using the Create network paths option. As a result, the lines and points locations are ready for the next steps in creating the hydrologic network.

The Identify Bluespot Features geoprocessing tool opens.

• For Input DTM Raster Dataset, browse to DHyMb located in the Inputs.gdb geodatabase.
• For Input Minimum Bluespot Depth, type 0.05 (meters).
• For Input Minimum Bluespot Volume, type 1 (cubic meter).
• For Output Workspace, specify your Outputs file geodatabase.

Next, you will explore the model outputs.

Identify model outputs such as Bluespots, PourPoints, BluespotWatersheds, and BSDepths.

• BuildingsNoGaps
• PourPoints
• Streams
• Bluespots
• BluespotWatersheds

In addition, notice that some buildings seem to be located critically within or next to bluespots—especially if located where outdoor staircases lead into their basements.

The symbology for the BSDepths layer does not clearly allow you to observe a bluespot’s depth when filled to their pour points. Next, you will modify the symbology of the layer to make this easier to observe.

Notice the current layer symbology is a stretch applying grayscale values for the depth range from 0 to 7.3 meters.

The default classification method is the Natural Breaks (Jenks) method with 5 classes. For this layer, this method will allow you to easily observe bluespot depths.

Different bluespot depths are now easier to differentiate; some of them are very deep—up to approximately 7 meters—and others are quite shallow.

Notice that when a bluespot is filled to its pour point, it is visible. In addition, the depth map includes all minor bluespots that were ignored by the threshold setting because they were shallower than 0.05 meters (5 cm) and had volumes < 1 m³. So, only the bluespots having a red pour point are preserved.

The table indicates that the study area has 3,388 preserved bluespots.

• There is a unique BluespotID for each feature
• Notice the asterisk after some field names indicating that the field is indexed to speed up performance when establishing joins later
• The maximum depth (in meters)
• The capacity (in m³) when filled to the pour point level
• The local contributing watershed area (in m²)
• The FillUp value (in mm), which is the ratio of a bluespot’s capacity in m³ versus its watershed area in m² multiplied by 1,000. This value indicates the amount of precipitation required to fill up a bluespot to its pour point level when assuming 100 percent surface runoff and no contributions from upstream spillovers.

Meanwhile, this ratio may not be as significant, as bluespots having small capacities but relatively large local watersheds (such as the first occurrence in the table) will fill up even at very small rain events. Which is, of course, only the case if the surface is totally impervious (which is part of the assumptions when using this screening method).

Notice its FillUp value is 1716.978506 mm and it has a very high Capacity (412927.286978) versus its local WatershedArea (240496.48).

This means that this bluespot will never get filled up by water coming from only its local watershed, since rainstorms of such magnitude (fortunately) never occur in Denmark. Meanwhile, this bluespot may have contributions from other upstream bluespots spilling over, and that is what you are going to investigate later.

The PourPoints layer contains attribute fields joined from the Bluespots layer.

• Each Pour Point is identified by its BluespotID.
• The Z field has information on the pour points elevation.
• The Capacity (in m³) of the bluespot.
• The pour points contributing WatershedArea (in m²)
• The FillUp value (in mm) is the ratio of a bluespots capacity in m³ versus its watershed area in m² multiplied by 1,000.

Since the PourPoints layer carries the necessary quantitative information needed to model the downstream spillovers from the bluespots, you will use this layer to model the downstream spillovers from the bluespots at various rain events.

Now the stream flow directions should be easy to inspect on the map. Arrows indicate the downstream flow from a bluespot to the next lower lying bluespot.

While reviewing this map subset, imagine the complexity of the total stream network, and how tedious it may be to calculate the accumulated spillover from one bluespot to the next ones downstream for the entire study area.

• For Input PourPoints Feature Layer, choose PourPoints.
• For Input Streams Feature Layer, choose Streams.
• For Output Workspace, choose your Outputs.gdb geodatabase.

In the Create Nodes and Edges tool, you may want to display the Results messages.

Note that there are still 3,388 true pour points, but 2,007 new junction nodes were added to the end of the attribute table.

When building the hydrologic network, the junction nodes are simply added to the pour point nodes in a Nodes feature class, where the node IDs match the corresponding BluespotIDs.

Use the Inspection of nodes bookmark to zoom to the extent of several small bluespots. If necessary, zoom in closer to inspect.

The red nodes are pour point nodes and the green ones are junction nodes located where streams intersect.

Notice how one stream in the south originates at the pour point with BluespotID 1116, and another stream in the west originates at the pour point with BluespotID 1102. Both streams end where they merge at the junction node with BluespotID 3808. From there, a new stream begins and ends at the pour point with BluespotID 1096, where a new stream begins and ends at the pour point with BluespotID 1079.

Ensure the Streams, PourPoints, and Nodes are selected.

As you scroll records in the table, you may notice the missing values <Null> for MaxDepth, Capacity, WatershedArea, and FillUp for the junction nodes. You will explore these later.

Notice that the Streams attributes include a BluespotID_start and BluespotID_end field.

You can now observe the sequence of flow from BluespotID 1116, and 1102, into 3808 and from 3808 into 1096 and then 1079.

Now the data modeling is complete. Before continuing, review the one-to-one relationships established in the model.

The only parameters required are the names of the Input Nodes Feature Layer or Feature Class that you created in the previous modeling steps and the Input Rain Event Start Value, the End Value, and the Increment Value in millimeters.

• For Input Nodes Feature Layer, choose Nodes.
• For Input Rain Event Start Value (mm), type 20.
• For Input Rain Event End Value (mm), type 100.
• For Input Rain Event Increment Value (mm), type 10.

The RainVolume20mm field contains the rain volume entering a bluespot from its local catchment (watershed) at a specific rain event. When looking at the table, the RainVolume20mm is calculated as 20 mm of precipitation times the WatershedArea 896.16 m² = 17.9232 m³.

During the execution, you can monitor the output messages in the View Details pane (see below). Notice the very fast processing time.

The table contains four fields for each of the nine rain events examined, in other words, 36 fields.

The pop-up appears and it is easier to review the four fields generated for each of the nine rain events examined, in other words, 36 fields.

• For Field Name, choose BluespotID.
• For Operator, choose includes the value(s).
• For Values, type 1079,1096,1102,1116,3808.

The table updates and filters to show only selected records. If necessary, expand the table to review additional fields.

This table contains four fields for each of the rain events examined. For each of them, the balance for each bluespot based on values stored in the Nodes feature class can be derived as follows:

SpillOverOut = SpillOverIn + RainVolume – Capacity

So, for the 20 mm rain event, you need to consider the balance obtained from these fields:

SpillOverOut20mm = SpillOverIn20mm + RainVolume20mm – Capacity

To get the results right, you start at the beginning of a local network and look at all values rounded to three decimals—for example, at the node with BluespotID 1116 that has a Capacity of 1.894 m³.

At a 20 mm rain event, the incoming rain volume from its surrounding local watershed saved in the RainVolume20mm field is 16.141 m³. As this bluespot has no upstream contribution, the SpillOverIn20mm field is 0, so the SpillOverOut20mm field is calculated as 16.141 - 1.894 m³ = 14.247 m³.

Next, explore the balance for the node with BluespotID 1102. The balance here sums to 10.700 m³.

Now, as the spillover for the nodes with BluespotID 1102 and 1116 merges at the junction node with BluespotID 3808, the junction node’s balance is totaled from a sum of the incoming spillovers from the nodes with BluespotID 1102 and 1116, only, as this node doesn’t have any contributing watershed. Thus, the balance at BluespotID 3808 is 24.947 m³.

Next, the balance at the node with BluespotID 1096 is 24.947 + 6 – 3.205 m³ = 27.742 m³.

This example demonstrates how the balances are worked out hierarchically, first considering the farthest upstream bluespots. Then the calculations go deeper and deeper into the network until the calculations end at the drainage basin’s terminal point located at the outlet into the sea.

Notice that several bluespots don’t produce spillovers because of their large capacities.

The Filled Events tool requires the Nodes feature class as input.

The values represent the rain event at which the bluespot starts spilling over.

Notice that several bluespots are assigned a field value of -1. These bluespots have a huge capacity and do not fill up, even at the at the most extreme rain events investigated.

Before you continue, you will add a new map to your project and copy only the necessary layers needed for interpretation.

A new map named Map1 is added to the project. You will use this map to investigate and explore the consequences of storm water in the Charlottenlund study area.

The selected layers are duplicated in the Investigate Stormwater map.

This will ensure point and line layers are displayed above polygon layers.

Labels are now placed at the lower right of the node point and are easily readable.

Notice that both the Bluespots and the BluespotWatersheds features have matching IDs = 1116.

Notice that the selected stream starts at the bluespot with ID value 1116 and ends at the bluespot with ID value 3808.

Since the BluespotID values are available and serve as a key field, you can transfer results in the Nodes layer to the Bluespots and Streams layers through joins.

This will remove node labels from the map.

The Bluespots layer is duplicated in the Contents pane. Next, you will rename the copy of the layer.

• For Input Table, verify FilledEvents is selected.
• For Input Join Field, choose BluespotID.
• For Join Table, choose Nodes,
• For Join Table Field, choose BluespotID.

Notice the addition of the storm water event modeling attributes originally contained in the Nodes layer.

The Symbology pane appears.

All filled events or bluespots are now symbolized according to the rain event where the incoming water volume exceeds a bluespot’s capacity, so the color represents the rain event at which a spillover begins to take place.

This also means that the bluespots spilling over have reached their maximum filled horizontal extent. All sinks symbolized in gray have such high capacities that they are not yet filled and will not spill over even at the biggest rain event modeled—in this case, the 100 mm event.

• For Category or Date, choose FilledEventmm.
• For Aggregation, verify Count is selected.

The chart shows the number of bluespots that are filled at a specific rain event (in mm) and start spilling over. Notice that the number of bluespots versus the FilledEventmm decreases asymptotically with increasing rain events.

On the map, notice the large bluespot located east in the study area will not spill over at a 100 mm rain event, either. But looking into its attributes, notice its ActualVolume100mm being 44,845 m³, which is very close to its capacity (45,150 m³). So, presumably, this bluespot will be filled entirely and start spilling over at a 110 mm event.

When clicking the 20 mm bar chart column, it shows that 2,689 bluespots are filled at that event. Simultaneously, the matching bluespots are highlighted on the map. Typically, such bluespots are shallow, have small volumes, and are found in gardens and along infrastructures.

The map and chart identify 126 bluespots. When zooming in to them, most are located as first order bluespots with FillUp values > 100 mm. Meanwhile, some are also located downstream, but have such large capacities that they don’t spill over even at the maximum rain event investigated.

Notice its capacity of 412,927 m³. When reviewing its attributes, notice that for the 100 mm event, the field named ActualVolume100mm shows that the actual stored water volume is 32,372 m³. Thus, buildings located along this depression’s upper rim will never flood.

• For Category or Date, choose FilledEventmm.
• For Aggregation, choose Sum.
• For Numeric Fields, click +Select and choose Capacity and click Apply.

In the chart, notice the bluespots not being filled up at a 100 mm event (still represented by a value of -1) make up the largest summed volume.

On the map, note that two large bluespots dominate the 70 mm category.

To estimate how many buildings might be affected during a rainstorm, a selection can be set up for features in the BuildingsNoGaps layer that intersects the filled bluespots at the 70 mm event.

• For Input Rows, choose FilledEvents.
• In the Expression builder, for Field Name, choose FilledEventmm.
• For Operator, choose is less than or equal to.
• For Value, choose 70.

As a result, 3,322 bluespots are selected.

Next, use the Select By Location tool to identify the BuildingsNoGaps features that overlap the selected filled bluespots.

• For Input Features, choose BuildingsNoGaps.
• For Relationship, choose Intersect.
• For Selecting Features, choose FilledEvents.

As a result, 3,126 buildings are selected. This means that 3,126 buildings are affected by bluespots, some more than others depending on their intersecting locations.

What can we learn about bluespots and buildings?

This analysis raises several interesting questions concerning a building’s flood risk due to its bluespot adjacency—especially if the bluespot is located next to a building where a staircase leads into the basement. In this case, the flood damage may be more severe than for similarly located buildings without basements. If information about staircase locations were available, the flood risk might be depicted much more accurately when combined with the bluespots’ depths around the staircases.

Another consideration is that buildings located completely within filled bluespots may be at higher risk than others. However, when selecting by location to identify buildings located completely within bluespots, no matches are returned, even if a visual inspection identifies such occurrences. Why is that?

• First, consider that all building footprint polygons have been converted from a vector to a raster format and are burned onto the DTM enforcing runoff around them. Logically, this means that we will never identify bluespots inside buildings.
• Second, the derived bluespot areas were originally a raster representation, converted to vector polygons without feature simplification applied.
• In combination, this means that when Select By Location is used on the Bluespots (or FilledEvents) layer using the Intersect spatial relationship, the vectorized bluespots are matched with the BuidlingsNoGaps vector footprints. Thus, overlaps are identified, as shown in the figure below, unless a building is oriented perfectly along the elevation model’s cells. If that is the case, no matches are found. In addition, if Select By Location is used on buildings located Completely Within bluespots, as mentioned before, no matches are identified.

• No matches identified is a limitation related to the data representations, and what was prioritized for the analysis: In this example, the location of buildings and the volumes that they occupy were prioritized over the ability to identify buildings located completely within filled bluespots.
• If you did want to select buildings located completely within bluespots, everything must be remodeled using a DTM without buildings incorporated. This would cause bluespot volumes to be inaccurate, as building volumes are not accounted for affecting bluespot capacity and spillover estimations.
• An additional consideration is that if you try to examine which buildings intersect bluespots at various filled events, some large buildings may intersect bluespots having different filled events. The example below shows a building affecting 20-, 50- and 60-mm events. In this situation, when deriving statistics to count the number of buildings located next to bluespots filled up at various events, the same building may be counted more than once. In these cases, the building should only be counted with the lowest filled event value, which in the example illustrated is 20 mm.

You also need to consider flood risks to buildings located in bluespots at rain events only causing a partial fill-up. In these cases, depth/volume relationships must be considered for the bluespots in question to estimate the water levels matching specific and actual bluespot volumes. As a result, a depth/volume relationship curve must be established for each bluespot incremented, for example, in steps of 5 centimeters depending on the spatial resolution and accuracy of the DTM used. In this way, water depth can be estimated at a given volume.

And lastly, the bluespot’s horizontal and partly filled (shrunken) extent can be depicted and buildings identified that intersected at that level. This analysis requires an additional set of tools that are not currently included in the model but are in consideration for future enhancements to the workflow.

The FilledEvents layer is already joined with the Nodes layer; copying the layer will preserve the join.

What did we learn about 60mm bluespot spillovers?

In general, the 60mm spillover is limited in the outskirts of the drainage basin, but as the bluespots overall don’t have capacities that can match the incoming volumes at high rain events, the runoff increases dramatically.

It seems like the volumes in m³ going downstream reach quite high values, especially in the lower central part of the study area next to or overlapping some vital infrastructure.

This is, indeed, critical along the roads toward the southwest that serve a large regional hospital (Gentofte Sygehus).

• For Input Table, verify StreamVolumes is selected.
• For Input Join Field, choose BluespotID_start.
• For Join Table, choose Nodes.
• For Join Table Field, choose BluespotID.

Notice that the class intervals for the StreamVolumes layer are not the same as for the SpillOver60mm layer, although both layers are symbolized using the same SpillOverOut60mm attribute values.

In the StreamVolumes layer, pseudo nodes located where streams merge are included in the join of nodes to streams, so, as streams merge at pseudo nodes, the spillover values increase. However, pseudo nodes located where streams merge are not included in the join of node attributes to bluespots and thus the spillover volumes do not increase.

• For 1937, type 2476.
• For 8072, type 9806.
• For 20684, type 26864.
• For 36859, type 57074.
• Do not change 67775.

Next, exclude streams not carrying any spillover.

Where SpillOverOut60mm is equal to 0

The map updates and streams are symbolized based on the spillover volume at the 60 mm event.

Where SpillOverOut60mm is equal to 0

What did we learn about stream spillovers?

Although this method doesn’t show the actual stream widths, it may provide a better overview of where the biggest water transport corridors are located in the study area. To map the actual widths and depths of the water corridors, kinematic wave equations may be considered, however, this is not currently addressed in the model.

Notice that if a stream’s width abruptly decreases downstream, it is due to a local retention of the runoff in a bluespot having a high capacity, thus reducing the downstream volumes leaving it.

Also, notice how streams in many places are identified along the roads' curbsides, as shown below.

When using a DTM of high precision, a road's centerline is sometimes visible as a drainage divide found along it, forcing the runoff onto the curbsides. That can be verified by swiping the BluespotWatersheds layer on/off when displayed on top of a road map.

The selected layer is added to the Investigate Stormwater map. Turn off all other layers.

• For 0.40422, type 0.25.
• For 1.126043, type 0.5.
• For 2.367577, type 1.
• For 4.330934, type 2.
• Do not change 7.362587.

The map updates and displays bluespot depths in meters.

What can we learn about bluespots' depths?

Apparently, several buildings are located critically within the deepest lying parts of the bluespots in the southeastern part of the drainage basin near Hellerup, not far from the basin’s outlet into the sea next to Charlottenlund Fort.

Note that all depths represent bluespots filled to their maximum levels (~ the bluespots’ pour point levels), so bluespot depths should always be compared with the FilledEvent layer.

According to the FilledEvent layer, this very large bluespot is filled up at a 70 mm event, and since many buildings are located in the bottom of the bluespot at depths around 2 meters, this area is likely to experience severe flooding problems, even at lower rain events.

Among emergency planners, such a map is popular and named a 'rubber boot index map,' as it tells the teams where to use rubber boots, waders, or sail around in dinghies when rescuing citizens.

Within the study area, classic flooding locations during rainstorms are located at underpasses where roads dip underneath railroad bridges. For an example, review the location at the bookmark named Bernstorffsvej Station.

When filled to its pour point level at a filled event of just 30 mm, this bluespot has a maximum depth of a staggering 3.25 meters. Many nonlocal drivers are stuck there during rainstorms when they underestimate its depth.

If you checked the original Streams layer, you will notice that several streams contribute to this major bluespot.

In reviewing the nodes and the pour points in the bluespot’s upper right corner, you will observe that its actual capacity is 15,388 m³, and at a 30 mm event, it produces a spillover of 8,908 m³. When the spillover is this high, it may be dangerous to even walk or drive here because of possible strong currents.

As you can see, the large spillover continues toward the east into a nearby bluespot and then farther eastward until the water moves toward the north along the infrastructure.

In the Change settings for the current project pane, review the current Home folder, Default geodatabase, and Default toolbox settings for the project.

The Outputs geodatabase stores output data generated by the modeling tools.

You may copy or import the feature classes into the Inputs geodatabase, or add a folder connection to an existing geodatabase containing your input data and DTM. Copying a feature class may only be done between geodatabases; if your data is in a non-geodatabase format, you would need to import instead of copying.

Your model data may include the following:

• DTM
• Building footprints
• Roads
• Railroads

These feature classes represent your processed data to be used in the tools and model to predict the flooding consequences from storm water events.

Review the model steps; when done, close and do not save changes.

The model executes, and a layer named BasinPolygons is added to the Bluespot Model Map.

After extracting the DTM for the basin applicable to your study area, you are ready to execute the Identify bluespot features model. However, you should consider incorporating buildings into your DTM to enable flows around structures and to eliminate the buildings' volumes if they are located within a bluespot. If you choose not to incorporate structures, you may continue by executing the model.

The Add Buildings to DTM model executes and the resulting DTM now incorporates building footprints that have slivers between adjacent buildings eliminated and gaps inside closed building complexes filled.

The illustration below shows the incorporation of structures into the Charlottenlund drainage basin DTM.

How the Burn Flowlines Onto DTM model works

For all line features, the model locates the lines' endpoints, and assigns elevation values to them extracted from the DTM.

For each line feature, the two endpoints' z-values are averaged, joined onto the line feature attribute table, and used as the value to be assigned to the cells during a line to raster conversion. The output is a temporary raster that is stamped onto the DTM.

When done, water flows are enabled at the approximated z-level through one cell wide hydrologic flow paths.