Abstract
Covariates derived from 1200 GMT tropospheric soundings taken for 13 June at Dodge City, Kansas, were compared with 3 h rainfall within the surrounding area. Correlation magnitudes decreased beginning with the first period for which precipitation lagged the soundings. There were larger decreases thereafter. It is suggested that the correlation magnitudes decrease because the measured environment is not representative of the environment within distant precipitation-producing migratory disturbances. This implies that atmospheric measurements should be taken simultaneously with the initiation of rainfall if maximum correlative power is to be attained.
Spatial analyses are presented as an alternative to frequent observations so that migratory precipitation-producing disturbances can be detected prior to the on-site rainfall. Covariates obtained from spatial fields gave higher correlations with 3 h rainfall than did the covariates obtained from single soundings.
