1: \begin{abstract}
2: \noindent
3: In [C.~W.~Gear, T.~J.~Kaper, I.~G.~Kevrekidis, and A.~Zagaris,
4: Projecting to a Slow Manifold:
5: Singularly Perturbed Systems and Legacy Codes,
6: {\it SIAM J. Appl. Dyn. Syst.}
7: \textbf{4} (2005) 711--732],
8: we developed
9: a class of iterative algorithms
10: within the context
11: of equation-free methods
12: to approximate
13: low-dimensional,
14: attracting,
15: slow manifolds
16: in systems
17: of differential equations
18: with multiple time scales.
19: %
20: For user-specified values
21: of a finite number
22: of the observables,
23: the $m-$th member
24: of the class
25: of algorithms
26: ($m = 0, 1, \ldots$)
27: finds iteratively
28: an approximation
29: of the appropriate zero
30: of the $(m+1)-$st time derivative
31: of the remaining variables
32: and
33: uses this root
34: to approximate the location
35: of the point
36: on the slow manifold
37: corresponding to these values
38: of the observables.
39: %
40: This article
41: is the first
42: of two articles
43: in which
44: the accuracy and convergence
45: of the iterative algorithms
46: are analyzed.
47: %
48: Here,
49: we work directly
50: with explicit fast--slow systems,
51: in which there is
52: an explicit small parameter,
53: $\eps$,
54: measuring the separation
55: of time scales.
56: %
57: We show that,
58: for each
59: $m = 0, 1, \ldots$,
60: the fixed point
61: of the iterative algorithm
62: approximates the slow manifold
63: up to and including
64: terms of ${\mathcal O}(\eps^m)$.
65: %
66: Moreover,
67: for each $m$,
68: we identify explicitly
69: the conditions
70: under which
71: the $m-$th iterative algorithm
72: converges to this fixed point.
73: Finally,
74: we show that
75: when
76: the iteration
77: is unstable
78: (or
79: converges slowly)
80: it may be stabilized
81: (or
82: its convergence
83: may be accelerated)
84: by application
85: of the Recursive Projection Method.
86: Alternatively,
87: the Newton--Krylov
88: Generalized Minimal Residual Method
89: may be used.
90: %
91: In the subsequent article,
92: we will consider
93: the accuracy and convergence
94: of the iterative algorithms
95: for a broader class
96: of systems---in which
97: there need not be
98: an explicit
99: small parameter---to which
100: the algorithms also apply.
101: \end{abstract}
102: