A self-balancing CUSUM approach for the efficient allocation of resources during post-disaster debris disposal operations

Hurricane Katrina and other recent disasters have underscored the challenges related to disaster-generated debris disposal. During Katrina, extraordinary amounts of debris, far exceeding typical annual amounts of solid waste, were almost instantaneously deposited across a three-state area. Collection and disposal of disaster debris is an enormous task. Although the locations and amounts of debris can be easily summarized after recovery activities have been completed, they are uncertain and difficult to estimate in real time. Inaccurate estimates can result in inequitable allocation of disposal resources, increased costs, prolonged recovery, and increased social, political, and economic unrest. This paper uses prospective statistical process control methods to achieve equity in allocating debris disposal resources. These methods enable the detection of emerging debris collection patterns in real time as debris information becomes available during disposal operations. Using the self-starting CUSUM method proposed by Hawkins (Statistician 36:299–315, 1987) as a foundation, we develop a self-balancing approach for debris cleanup operations and evaluate its performance using data from a 2003 Atlantic hurricane.