other disciplines. This is true of many questions in sustainability science. In the
case of the hurricane, for example, there are many more questions that need to
be asked, increasingly complex and subtle questions. A key role of mathematical
scientists, collaborating with others, is to help pose the right questions. For
example: What happens after the evacuation? Or suppose that the hurricane
occurs at the time of an epidemic when people are in quarantine; how does this
change our response? Also, how much does it cost to repair the damage? Is the
economy of the region destroyed? What are the indirect economic effects of the
disaster and how can they be measured? And is it a good idea to repair the
damage or should the center of the city be moved over the long term? Could we
have prevented some of the damage by building dikes? Should we invest in
dikes for the future, considering that hurricanes are likely to become more
frequent and stronger in the future? What if there are several hurricanes in a
period of a few years? These are just some of the questions we can ask. For
each of the examples given in this report, we could ask many similar questions.
Operations research and discrete mathematics have long studied
questions like those about evacuation, but answering such questions pushes
existing tools beyond their capabilities. For example, in an evacuation, decisions
need to be made about how many doctors and nurses need to stay behind to
care for those who don’t evacuate, and where those medical personnel should be
assigned. This “job assignment problem” is a classical one in operations
research, but existing techniques don’t deal with uncertainty well. In a real
evacuation, uncertainty is huge: How many doctors and nurses will be willing to
put themselves at risk by staying behind? How long will the city remain
inundated? How many people will need care, and what will their medical needs
be? Planning in the face of this uncertainty will require new tools in the field
known as “stochastic optimization” (optimization under uncertainty/randomness).
Similar questions surround stockpiling supplies at evacuation sites. Inventory
planning has long been studied in operations research, but existing methods
don’t deal well with uncertainty of the kind we might experience.
Human well-being includes adequate food, housing, and water; good
health; a secure and pleasant environment (one protected against natural
disasters as well as threatening changes in the climate, rising sea level, etc.);
and a prosperous economy so that people have jobs. The hurricane scenarios
illustrate how human well-being, in all these senses, is intimately connected to
the health of our ecosystems. We have learned that our decisions enormously