Module 8 - Damage Assessment

This was our last lab before our final project! It's been an intense semester. We wrapped things up by looking at coastal flooding again, but from the perspective of assessing damages done by a real natural disaster -- that of Hurricane Sandy's in October of 2012. Although Hurricane Sandy made landfall as a Category 1 hurricane on the coast of New Jersey, it was extraordinary in size, and thus created a lot of damage. For our lab assignment, we conducted a damage assessment by being placed in the seat of a GIS Technician working for a FEMA contractor in the immediate hours after the disaster. Below is my final map product depicting the path of Hurricane Sandy, the affected U.S. States, and intensity/strength of the cyclone during its track.

In addition, we also prepared data by carrying out a building damage assessment based on imagery from before and after the storm within a study area. To minimize user-input error, we created attribute domains in a geodatabase. Two important facts to remember about domains: 1. they are declarations of acceptable values, and 2. using domains help ensure data integrity by limiting the choice of values for a particular field. Below are screenshots of all the domains I created, along with their properties and assigned coded values:

Below is a screenshot of my damage assessment results, showing all the points I created, symbolized by the structural damage.

My process for identifying properties included zooming in close enough were it wasn’t too pixelated, and then using the Swipe Tool various times back-and-forth to determine the level of structural damage. I started out on the East coast, and worked my way westward because it was easier to determine the properties that were completely “Destroyed” versus the ones that seemed to be “Affected” or with “Minor Damages”. Personally, the decisions that were most difficult were determining the level of “Wind Damages”. I think this category is ambiguous. I had to Google “wind damage”, and the majority of the images I saw were of rooftops. So I based my judgment out of how clean the rooftops looked on the imagery. Still, I also consider debris to be a cause of wind damage. That’s why I had a hard time with this category. Another category I struggled with at first was determining the “Inundation” status. After going through all the parcels, I decided to just make all of the points as “Yes” to “Inundation”. The fact that roads and sidewalks were completely covered with sand makes me think that the entire place was under at one point. So technically, these properties are not under water at the moment, but there was a level of inundation, and so because of this ambiguity, I had some difficulty making decisions with this category as well. The only other layer that I think could have been useful in this analysis is a vegetation/landcover feature class showing plants, trees, shrubs, etc. I think this would have been helpful because if vegetation is destroyed from the velocity of wind, it can have a direct impact on structural damages that are on its way.

Furthermore, we examined the locations of the damaged properties relative to the coastline. By using a Proximity Analysis tool (Buffer Wizard), I was able to quickly ascertain the counts of structures within 100, 200, and 300 meters from the coastline. I've provided a screenshot to show my multiple buffer rings, my digitized coastline polyline, and my calculations summarized in a table, see below.

I created 120 points in my Structure Damage feature class. Out of 120 points, 0% are No Damage, 5% have Minor Damage, 16% are Affected, 29% are Destroyed, and 50% are Major Damage. Within 100m of distance from the coastline, 100% of buildings are damaged. Within 200m of distance from the coastline, 91% of buildings are damaged, and within 300m of distance from the coastline, 70% of buildings are damaged. Based on these calculations within this study area, we can extrapolate to say that this storm surge has caused major building damage, with the most being within 200m from the coastline all along the New Jersey southeastern coast, which has an approximate length of 100 miles. Based on the data we were provided of Hurricane Sandy’s track and location, Sandy’s center (eye) made landfall 50 miles south from this study area as a Category 1 Hurricane with winds of 90 miles per hour. New Jersey emergency responders should prepare to assist a coastal area of at least 18 squared miles or 11,930 acres large. I’m sure that insurance rate companies would use this information to increase premiums for buildings within 300m of the coastline.