332
D.G. Chen
11.5
11.0
ti10.5 I----++--t-I--F-----t------+----+--+-+---J:--i
CI)
10.0
9.5
1950
1960
1970
1980
1990
Year
Figure 17.1. Time series of the annual mean sea surface temperature (0C) at
Amphitrite Point, west coast of Vancouver Island. The horizontalline is the longterm time series average. Note year 1977 is the year for the transition to the
current warm climate regime.
In general, most of these environmental factors are intrinsic fuzzy terms and
there is no crisp and clear break point for the classification. Therefore a fuzzy
logic approach should lead to an improved SR analysis.
17.2.2
Fuzzy Stock-Recruitment Model
A fuzzy logic model is also known as a fuzzy inference system or fuzzy-rulebased system. Basically, any fuzzy logic model consists of three parts, wh ich are
the fuzzy membership functions, fuzzy decision rules, and the fuzzy reasoning.
Several types of fuzzy reasoning have been developed in the literature (Bandemer
and Gottwald 1995; Lee 1990). Following the tradition al SR model (17.3), a
fuzzy logic SR model (hereafter referred as Fuzzy-SR) is proposed in this paper to
model and classify the fish SR relationship. Without loss of generality, the
description of this Fuzzy-SR model is restricted to only two environmental
regimes, i.e. such as "Cool" and "Warm". The extension to any number ofregimes
D.G. Chen
11.5
11.0
ti10.5 I----++--t-I--F-----t------+----+--+-+---J:--i
CI)
10.0
9.5
1950
1960
1970
1980
1990
Year
Figure 17.1. Time series of the annual mean sea surface temperature (0C) at
Amphitrite Point, west coast of Vancouver Island. The horizontalline is the longterm time series average. Note year 1977 is the year for the transition to the
current warm climate regime.
In general, most of these environmental factors are intrinsic fuzzy terms and
there is no crisp and clear break point for the classification. Therefore a fuzzy
logic approach should lead to an improved SR analysis.
17.2.2
Fuzzy Stock-Recruitment Model
A fuzzy logic model is also known as a fuzzy inference system or fuzzy-rulebased system. Basically, any fuzzy logic model consists of three parts, wh ich are
the fuzzy membership functions, fuzzy decision rules, and the fuzzy reasoning.
Several types of fuzzy reasoning have been developed in the literature (Bandemer
and Gottwald 1995; Lee 1990). Following the tradition al SR model (17.3), a
fuzzy logic SR model (hereafter referred as Fuzzy-SR) is proposed in this paper to
model and classify the fish SR relationship. Without loss of generality, the
description of this Fuzzy-SR model is restricted to only two environmental
regimes, i.e. such as "Cool" and "Warm". The extension to any number ofregimes