Super Storm Sandy Damage Assessment

Here's a link to the Story Map I created about the products of this analysis.

After creating raster mosaics of the pre-storm and post-storm imagery, using Flicker and Swipe were an easy way to compare damage to structures.  Considering the proximity to the shoreline, if the swale and driveway of a property were washed away, I considered the property inundated.  I started with a scale of 1:1,600 and would swipe off and on the PostStorm imagery file.  Immediately after, I would zoom in to a property at a scale of 1:275 to determine any structural or wind damage.  Buildings with sheds that were blown away were considered wind damage and not structural damage.  If a building’s outline was changed, this was logged as either minor or major damage.  

Now that comparing is made easy, I created a new DOMAIN in the default GDB representing specific types of damage that are to be included in the assessment (inundation, structural damage, wind damage) as well as the type of structure.  Once the domain was built in the GDB, I created a point feature class representing the damage in a parcel, one point feature to be within each individual parcel.  Once a point was added to every parcel, I edited the attribute data through the attribute table.  By using Flicker and Swipe, I was able to approximate the type of damage to the selected point and assign appropriate attribute data for each point.

After digitizing the coastline, I created a buffer of 100m and 200m.  I created another buffer of the resulting feature classes of another 100m but did not include the original polygon to get the ranges of 100m-200m and 200m-300m.  After the buffers were created, I isolated the types of structural damage from the point features.  To do this, I used SELECT BY ATTRIBUTES from the Structure Damage feature class and different clauses to create a point layer for the No Damage, Affected, Minor Damage, Major Damage, and Destroyed points.  Using the new feature classes, I used SPATIAL JOIN on the OceanCountyParcel polygon layer to create four resulting feature classes that included a Join_Count.  This still resulted in a feature class including all the parcels, so I again used SELECT BY ATTRIBUTES to isolate the parcels with a Join_Count as 1 and exported the features to create a new feature class with only the necessary parcels included.

After the four categorized parcel layers are created, I ran SELECT BY LOCATION and isolated the parcels that were within each buffer by opening the attribute table once the process was ran.  I originally tried a SPATIAL JOIN with the Structure Damage point feature class created and the OceanCountyParcels layer, but the buffer would only include the points and not the polygon parcel layer the points were within.  To keep all the attribute data, I had to do this roundabout method to ensure that the buffer included the parcel associated with the type of the structural damage and not only the arbitrary point that represents the parcel.