You are here:
USING FISHER INFORMATION TO ASSESS THE RISK OF DYNAMIC REGIME CHANGES IN ECOLOGICAL SYSTEMS
Citation:
Mayer**, A, C. W. Pawlowski**, AND H C. Cabezas*. USING FISHER INFORMATION TO ASSESS THE RISK OF DYNAMIC REGIME CHANGES IN ECOLOGICAL SYSTEMS. Presented at INFORMS Symposium on Managing Risk in an Uncertain World, Evanston, IL, May 31, 2003.
Impact/Purpose:
To inform the public
Description:
The sustainable nature of particular dynamic regimes of ecosystems is an increasingly integral aspect of many ecological, economic, and social decisions. As ecosystems experience perturbations of varying regularity and intensity, they may either remain within the state space neighborhood of the current regime, or "flip" into the neighborhood of a regime with different characteristics. Developing indicators that can determine when these systems are at risk of a regime change would be highly desirable for proactive adaptive management decisions. Information theory has significantly improved our ability to quantify the organizational complexity inherent in systems in spite of imperfect observations or "signals" from the source system. Fisher Information (FI) is one of several metrics developed within the area of estimation theory. FI can be described in three ways: as a measure of the degree to which a parameter (or state of a system) can be estimated; as a measure of the relative amount of information that exists between different states of a system; and as a measure of the disorder or chaos of a system. Highly disordered, chaotic systems have a low probability of being observed in any one particular state, and therefore have low information. Conversely, systems that follow a regular or repeating trajectory have higher information. FI may be a very useful measure to apply to the state of the system in order to identify the degree to which a system is at risk of "flipping" into a different steady state.
We have developed an FI index for dynamic systems, and examined data collected from several types of ecological and social systems. Previous research has identified the presence of stable states and transitions between them in several time series. These systems span a range of scales (from single lakes to the global climate), and demonstrate slow and fast transitions between regimes. All of these datasets are noisy, and reflect several to many cycles that are out of phase and operate over a range of timescales, which make detecting regime transitions difficult when looking at all of the variables simultaneously. For these ecosystems, FI successfully combined all of these variables and identified the transitions between regimes in the past. These results indicate that FI may be able to indicate when systems currently in a stable dynamic regime are entering a transition phase. Humans may be able to reverse behavior or inputs into the system to prevent the system's flip into a less-desirable steady state (or continue the behavior if the resultant steady state is desirable, such as in ecosystem restoration efforts), if systems entering these transition phases could be detected early enough.