(A)ccurate estimation of the reaeration-rate coefficient (K-2) is extremely
important for waste-load allocation. Currently, available K-2 estimation e
quations generally yield poor estimates when applied to stream conditions d
ifferent from those for which the equations were derived because they were
derived from small databases composed of potentially highly inaccurate meas
urements. A large data set of K-2 measurements made with tracer-gas methods
was compiled from U.S. Geological Survey studies. This compilation include
d 493 reaches on 166 streams in 23 states. Careful screening to detect and
eliminate erroneous measurements reduced the data set to 371 measurements.
These measurements were divided into four subgroups on the basis of flow re
gime (channel control or pool and riffle) and stream scale (discharge great
er than or less than 0.556 m(3)/s). Multiple linear regression in logarithm
s was applied to relate K-2 to 12 stream hydraulic and water-quality charac
teristics. The resulting best-estimation equations had the form of semiempi
rical equations that included the rate of energy dissipation and discharge
or depth and width as variables. For equation verification, a data set of K
-2 measurements made with tracer-gas procedures by other agencies was compi
led from the literature. This compilation included 127 reaches on at least
24 streams in at least seven states. The standard error of estimate obtaine
d when applying the developed equations to the U.S. Geological Survey data
set ranged from 44 to 61%, whereas the standard error of estimate was 78& w
hen applied to the verification data set.