Over the past 30 years, ecologists have demonstrated the importance of flow
and temperature as primary variables in driving running water, riparian an
d floodplain ecosystems. As it is important to assess the size and timing o
f discharge variations in relation to those in temperature, a method is pro
posed that uses multivariate techniques to separately classify annual disch
arge and temperature regimes according to their 'shape' and 'magnitude', an
d which then combines the classifications. This paper: (i) describes a gene
rally applicable method, (ii) tests the method by applying it to riparian s
ystems on four British rivers using a 20-year record (1977-97) of flow and
air temperature; (iii) proposes a hydroecological interpretation of the cla
ssification; (iv) considers the degree to which the methodology might provi
de information to support the design of ecologically acceptable flow regime
s.
'Regimes' are defined for discharge and air temperature using monthly mean
data. The results of applying the classification procedure to four British
rivers indicates that the 'typical' regimes for each of the four catchments
are composite features produced by a small number of clearly defined annua
l types that reflect interannual variability in hydroclimatological conditi
ons. Annual discharge patterns are dominated by three 'shape' classes (acco
unting for 94% of the station years: class A, early (November) peak; class
B, intermediate (December-January) peak; and class C, late (March) peak) an
d one 'magnitude' class (70% of the station years fall into class 3, interm
ediate), with two subordinate 'magnitude' classes: low-flow years (18%) and
high flow years (12%). For air temperature, annual patterns are classified
evenly into three 'shape' and four 'magnitude' classes. It is argued that
this variety of flow-temperature patterns is important for sustaining ecosy
stem integrity and for establishing benchmark flow regimes and associated f
requencies to aid river management. Copyright (C) 2000 John Wiley & Sons, L
td.