Conceptualization

A working assumption of this study is that the most effective way to model recreational boat traffic is from the bottom up (discussed in section 2.3.2). There are a number of processes happening in a recreational boat traffic system at a number of different temporal and spatial scales, and what happens at the coarser levels of resolutions arguably are largely emergent properties of what happens at the finer spatial and temporal scales.

  

Figure 4.1: Scales in hierarchy of decisions in boating activities

A convenient way to conceptualize the structure of spatio-temporal hierarchy of boater decisions and boat traffic behavior is presented in figure 4.1. This illustration is similar to a hierarchy proposed by Holling (1992) for decisions made by wading birds in their habitats. At the finest level of resolution is the personal buffer around the boat, which spatially could be from a few meters to a few tens of meters around the boat, and temporally would involve things that are happening in a matter of minutes but typically less than an hour. These include safety issues (avoidance of collisions, adjusting speed to navigate between obstacles, or dealing with wakes or swells), or spontaneous acts that may temporarily interrupt the flow of a trip, such as stopping to greet a familiar face, observe wildlife, or receive a citation from a friendly law enforcement officer.

At a slightly coarser level of resolution is a trip, which would range spatially from a few hundred meters to a few tens of kilometers, and temporally from less than an hour to an entire day. A trip may contain a number of stops, and may or may not be planned around what happens at the personal buffer level. One can quantitatively describe a trip considering only things that happen at the trip level (time of departure, destination, path taken, speed, etc.), but description of the quality of the trip would be described in terms of what occurred at the personal buffer level.

The next level up for boater activities would be in terms of a cruise, which could be a few kilometers or hundreds of kilometers, and would involve a time span of a day up to several years. This is a level where the boater is making longer term commitments to overall activities in which s/he is participating. Individual trips are planned in the context of the larger cruise, and there is a decreasing significance of events at the lowest levels in behavior at that level: events that happen at the personal buffer level are more likely to disrupt a single trip than they are to disrupt an entire cruise itinerary.

At the coarsest level of resolution in boater decisions is residence, where the boater decides to keep their boat for an extended period of time. This is analogous to the home range concept in animal behavior or the idea of community in socioeconomic studies. This involves significant commitment (and possible financial investment) on the part of the boater to the area, in terms of finding long-term facilities and services for themselves and their boat(s). Residence time can be anywhere from a few days to several decades, and the decision to reside in the area would probably involve things that are within a few meters to several hundred kilometers radius of the location where the boat is kept. Again, a location of residence can be perfectly described quantitatively (location, channel access, etc.) without considering things at the lower levels, but the quality of the boater's residence would be determined largely by the quality of the trips and cruises that the boater makes from there. A single bad trip may not cause a boater to reject a marina or dock as a place to keep their boat, but a series of bad trips would probably cause the boater to conclude that the residence is a bad boating location.

The hierarchical organization of the system resembles representations in economic models presented in Lin et al. (1996). In terms of the socio-spatial interaction models described in Dendrinos and Sonis (1990), where

$$x_i(t+1) = \frac{A_i F_i [x_i (t)...]}{\sum A_i F_i }$$ (2)

where x_i is the i ^th state variable at t+1, F_i[x_i(t)...] is a function of item i over all x_i at time t, and A_i is a ``long term environmental variable'' that represents slower changing dynamic processes on item i at time t. The term A_i can be considered to represent processes that are happening at the higher levels of organization. In context of the gravity models described in Wilson and Bennett (1985) and the like, the parameters of the system in question can be seen as a product of processes taking place at lower levels of organization, which could be a source of the variability addressed in Fik (1997) that was mentioned on page of this study.

This study is considering boat traffic at the trafficshed and regional scales. Trafficsheds typically occupy an area of a few hundred meters or a few square kilometers, and the corresponding temporal dynamics (according to the relationship presented in figure 4.1) would be from less than an hour to the course of a day. Traffic at the regional scale (Sarasota Bay) would imply a similar time scale. The fundamental activity of consideration in this system will be the trips.

Processes at coarser resolutions (residence of boaters) will be considered as ``long term'' variables, and will be considered to be constant over the course of a single run of the simulation (i.e., boaters changing residence will not be considered part of the system). Activities at finer resolutions will be considered only in terms of how they affect the course of a trip.