The presence of autocorrelation in the analysis of a variable sampled seque
ntially at regular time intervals appears to be unknown to many agricultura
l meteorologists despite abundant documentation found in the traditional me
teorological and statistical literature. It follows that the statistical co
nsequences as well as methodological alternatives are also unknown. Through
an example using paired radiometer observations, this note discusses recog
nition of autocorrelation as well as the importance of testing ordinary lea
st squares regression parameters in the presence of autocorrelated residual
s. An autoregression example is presented as one alternative way to analyze
the given dataset. (C) 1999 Elsevier Science B.V. All rights reserved.