Determining Flood Risk

Spectrum™ Technology Platform comes with two templates that illustrate how to determine the relationship of an address to known flood zones. The two templates are FloodRiskAnalysis and FloodRiskDetailAnalysis. Both these templates take an address or intersection and determine how prone the location is to flooding. The difference is in the method of comparison. FloodRiskAnalysis locates all the nearby flood zones then performs a union operation on those flood zones, which creates a new geometry that encompasses the flood zones. The address is then compared to this single geometry and an overlap percentage is returned. FloodRiskDetailAnalysis, however, compares the address to all nearby flood zones individually, returning separate overlap percentages for each nearby flood zone and allowing for a more detailed analysis of risk.

Both FloodRiskAnalysis and FloodRiskDetailAnalysis use two other dataflows, which are also provided as dataflow templates:

Template names: FloodRiskAnalysis, FloodRiskDetailAnalysis

Sample input file name: None (template is a service dataflow)

Modules required: GeoConfidence Module, Enterprise Geocoding Module, Location Intelligence Module

Business Scenario

You work for an insurance company as an underwriter for homeowner policies. As part of the underwriting process you need to know how likely it is that a home will experience flooding, which means you need to determine how close the home is to a flood zone. You know the address of home and from that you need to determine the flood risk.

FloodRiskDetailAnalysis compares the location to each nearby flood zone individually, returning separate overlap percentages for each nearby flood zone. The following dataflow provides a solution to the business scenario:

In this dataflow, the input address is first geocoded, meaning the latitude/longitude coordinates of the address are determined. Then the location is converted to a geoconfidence shape, which represents the area around the address. Next the nearest flood zones are located. Then the location's shape is compared to each flood zone shape and the percentage overlap is determined. Finally any records with a percentage of less than one are discarded, leaving the more significant results to be returned. Using these percentages you could then determine the degree of flood risk associated with the location.

This dataflow template is available in Enterprise Designer. Go to File > New > Dataflow > From template and select FloodRiskAnalysis or FloodRiskDetailAnalysis

Each stage is described in detail below.


The input to this dataflow is an address. Since this dataflow is a service, the first stage is an input stage. Note that there is an ID field which is used to preserve the order of the records in the output. The records are sorted in ascending order based on the value in the ID field. So for example, the first record could have an ID of 1, the second and ID of 2, and so forth.

Geocode US Address

This stage determines the latitude/longitude coordinates of the address and returns the information needed to construct a geoconfidence surface. A geoconfidence surface is a buffered point, line, or polygon, that encompasses the immediate area around the location. (Note that the Geo Confidence check box on the Output Data tab must be checked in order for Geocode US Address to return this information.) The type of shape returned as the geoconfidence surface depends on the precision of the match:


The GeoConfidenceSurface stage creates a shape that represents the area immediately surrounding the address or intersection. The GeoConfidenceSurface stage is a service. To view it, right-click the stage and select Edit This Dataflow.

The first stage in this dataflow after input is Centroid. This stage converts the point represented by the GeoConfidenceLatitude and GeoConfidenceLongitude values to a point geometry. This is necessary because the operations performed later in the subflow require the point to be represented as a geometry datatype as opposed latitude/longitude strings.

The RouterGeoConfidence stage then directs each record to the appropriate path based on the type geoconfidence shape determined by Geocode US Address:

Note: To use this dataflow you must first specify a database in the ZIP stage. To do this, install the database, then define a database resource for the database in the Management Console. For instructions, see the Spectrum™ Technology Platform Administration Guide. Then select the database resource in the Spatial database field.

Find Nearest

The Find Nearest stage locates the closest flood plains to the geoconfidence surface. Before running this dataflow you must specify a spatial database containing the flood plain data. To do this, first define the database resource in Management Console, then select the database resource in the Spatial database field. For instructions on defining a database resource, see the Spectrum™ Technology Platform Administration Guide.

The Find Nearest stage is configured to search up to 1 mile for nearby flood plains, and will return up to 50 flood plains. (You can reduce this if you want to limit the number of nearby flood plains returned.)


The intersection stage compares the geoconfidence surface with the flood planes and determines the percentage of the geoconfidence surface that overlaps with each flood plain. For example if the geoconfidence surface is the yellow circle and the flood plain is the rectangle, the overlapping area would be:

The percentage returned would be the percentage of the circle represented by the shape shown under "Output".

Filter Overlap

The Filter Overlap stage removes records where there is an overlap of less than 1%. This can be modified as needed to filter out those intersections that are not significant enough to warrant your attention.


The Output stage returns a variety of information, including the address, percentage of the geoconfidence surface that overlaps with each flood plain, the number of nearby flood plains, the latitude/longitude of the address, and more.