Evaluation of New Sensors for Emergency Management
9
4
LAKELAND IMAGERY STUDY
Proxy hurricane damage
While an ideal test of the EMERGE capability would be to map areas during
a real disaster, no events causing roof damage or hurricanes struck the East Coast
of the U.S. during the time when this study was being conducted (October 2000
through September 2001). This necessitated the selection of a proxy location
where it would be possible to observe features similar to those from hurricane or
other wind damage that would result in the need for temporary roofing. To deter-
mine the number of houses that would receive temporary roofing, it is necessary
to determine both whether the roof has been damaged and the status of the under-
lying rafters. After a ground reconnaissance, it was determined that active
construction of new housing in the area of Lakeland, Florida, provided an accep-
table analog to hurricane-induced roof damage. In this study, imagery was taken
of sites with houses in all phases of construction, particularly houses with roof
rafters intact but without plywood sheathing, with plywood sheathing but without
tarpaper or shingle covering, and with tarpaper-covered roofs.
Location
Lakeland is located about 30 miles east-northeast of Tampa, Florida, in the
vicinity of several freshwater lakes (Figure 1). Within the boundary of the city of
Lakeland, two areas were chosen for evaluation of EMERGE imagery as a means
of detecting roof damage and the status of roof rafters on partially completed
buildings. Data were collected on 11 February 2001 over area 1 (4 sq mi) and on
12 February 2001 over area 2 (4 sq mi) (Figure 2).
Fifty-three frames (35 frames at 1-ft GSD, 12 frames at 2-ft GSD, and 6
frames at 3-ft GSD) were collected over area 1 on 11 February and over area 2 on
12 February. The demonstration exercise began each morning with acquisition of
the data when the sky had cleared and ended by 5:00 p.m. with completely post-
processed orthorectified image frames. From examination of this data, it was
determined that imagery with resolution higher than that acquired was needed for
reliable visual observation of individual roof rafters. Data were again taken over
areas 1 and 2: on 28 March, 198 frames at 8-in. GSD and 70 at 1-ft GSD were
collected, and on 30 March, 18 at 2-ft GSD were collected. The same construc-
tion sites were imaged as before, but some houses had been completed and some
Previous Page
