By Scott Cox, product manager, Sanborn (www.sanborn.com), Colorado Springs, Colo.

For more than a century, insurers have recognized the value of geographic information for risk management. One of the first providers of such information was The Sanborn Map Company Inc., Colorado Springs, Colo. During the late 1800s, the company began offering detailed fire insurance maps, which underwriters used to help determine risks associated with urban properties.

The Sanborn Fire Insurance Map product line (commonly referred to as Sanborn Maps) came to prominence in 1871 when the Great Chicago Fire wiped out some $200 million dollars in property—about one-third of the valuation of the entire city—including 18,000 buildings. But buildings and property weren’t the only things destroyed. Unable to meet their claim obligations, many local insurance companies also went up in smoke. The Great Chicago fire made it painfully obvious to insurers that they needed better risk-evaluation tools.

Like the Great Chicago Fire, the destruction of the World Trade Center by terrorists highlighted the relationship between geographic information and risk management. The attack once again made it clear to insurers that the industry would benefit greatly from better risk-evaluation tools.

GIS in Risk Management
There are many ways insurance companies use maps and GIS to manage risk. The Sanborn Maps, which include data such as building construction material, the presence of sprinkler systems, the structure’s proximity to fire hydrants, etc., have been used by property and casualty underwriters to help them better determine a building’s risk. Insurers of other types of hazards also benefit from geographic analysis. For example, earthquake zones are mapped to help determine the amount of risk per region; hurricane paths are charted to determine the likelihood of a storm in a geographic area; and flood maps show spatially what level of flood insurance is required.

In the past, insurance companies would chart insured locations with push pins and a wall map. GIS technology has revolutionized this system, but the desired result remains the same: By reviewing data spatially, insurers can determine their amount of risk, which they calculate by multiplying their exposure (the amount of insurance written for a geographic area) by the hazard level in that area (the likelihood of a disaster). As an added bonus, the geographic data used to determine exposure are invaluable for disaster response. Since 2001, a new type of hazard concerns insurers: terrorism.

Data Acquisition
Sanborn began the data-gathering process by flying aerial photography over 26 U.S. cities. Obtaining permission for the program from local authorities wasn’t always easy. After 9/11 it was particularly difficult to gain permission to fly over Washington, D.C. In the end, after many delays, Washington officials allowed the flyover, albeit with stipulations. Washington was flown at 9,750 feet above ground level—twice the normal flying height for the project.

The aerial photography for all the cities except Washington was rectified to .5-foot-resolution orthophoto imagery and met 1:100-scale national map accuracy standards. (The Washington orthophotos were 1-foot resolution.) Once the photography was captured, stereo models were set, and building outlines and heights were compiled. Building heights, along with building dimensions, are important to insurers, because such information allows them to calculate approximate square footage.

Building Attribution
After the buildings were captured in 3-D, Sanborn added intelligence to the polygons by designing a database with fields for primary addresses and building names. Hard-copy Sanborn Maps were scanned and georeferenced, and the data were transferred from the annotated maps to the database fields. Such data included number of stories, occupancy type, construction type, year of construction and fireproofing information. The database was linked to the building polygons with a latitude and longitude coordinate taken from the centroid of the building.

Sanborn Maps originally were made for more than 12,000 U.S. cities and towns. During the last several decades, however, only select major cities were updated, and maps for several of the cities in this project were outdated. When current maps weren’t available for a particular city, surveyors went to the field to collect the relevant data. To populate the data fields, the surveyors were sent out with GeoXT hand-held computers from Trimble Navigation Ltd., Sunnyvale, Calif., and ArcPad software from ESRI Inc., Redlands, Calif. Cities were subdivided into manageable survey areas for each day, and the files were sent by e-mail back to the office for quality control each night. The surveyors also photographed buildings as necessary. Once a city was surveyed and passed quality control, the surveyors moved to the next city.

Building names—separate from the address, but still usable for geocoding— can be subject to unusual circumstances. For example, building names can be on more than one building (e.g., “Greenway Apartments”), and some buildings have multiple names (such as the Aon Center in Chicago, which is also known as the Amoco Building and the Standard Oil Building).

Because Sanborn Maps include interior firewalls, they’re useful for subdividing most buildings when a division isn’t apparent from aerial photography. However, there were cases where the Sanborn Maps didn’t contain enough information for technicians to subdivide large building complexes. To refine the data, Sanborn researched buildings on the Internet, and in some cases telephoned locations to resolve questions regarding addresses or names.

Some complexes defy standardization. The Prudential Center in Boston is one such building. In the end, RMS and Sanborn decided to depict the 52-story Prudential Tower as a separate building from the surrounding three-story Prudential Center. Because both buildings share an address, it was necessary to show one building with a building name, but without an address. In a complex case such as this, users must view a building “in context.” The orthophoto imagery is invaluable for understanding how buildings are connected, and how the connection may affect risk.

All addresses were checked against geocoding software to ensure they fell within valid U.S. Postal Service ranges. The final datasets were integrated into the RMS software, which includes a map and orthophoto imagery viewer that allows insurers to review the data when analyzing a building.

The datasets include detailed 3-D models, though insurers aren’t currently using such data. However, 3-D visualization is possible, and software has been refined to make 3-D interfaces practical for all geographic data users.

With building-level geocoding, insurers have a more robust and accurate analysis of their exposure concentrations in urban areas. This capability is particularly important when analyzing risk exposure from terrorism.

“Historically, many insurers used Sanborn Maps to track their exposure concentrations, using pins and paper versions of the maps to do so,” said Tuttle. “Now, using RMS products and Sanborn data, insurers can do the same thing electronically.”
 

  See  Featured Images
 
  Subscribe to Earth Imaging Journal

  Go to Home Page