WMOD00284 - Laserfiche WebLink

Home Browse Search

From the precipitation records at the network gauge sites it was evident that a storm almost always consisted of more than one precipitation period (hereafter called a precipitation episode). In fact, as will be shown, it was the norm that several shorter duration precipitation episodes constituted the entire storm. Precipitation periods at each of the seven gauge locations were examined to determine how long a typical precipitation episode might be expected to last. In this context, an episode was defined as follows: An episode began and continued whenever hourly precipitation amounts of any magnitude were recorded, including chart readings at the lowest resolution limit of the gauge. The episode continued as long as the precipitation was continuously detected, with the exception that up to a 2 hour interval without precipitation was allowed. That is, if precipitation resumed after no more than a 2 hour break the episode was considered continuous. This interval was adopted so that a comparison of the episode duration in the Arizona storms could be made with those episodes found in an analysis of Colorado's Grand Mesa storms (Super et. al., 1986). They developed this episode definition after a detailed examination of the hourly data from the Grand Mesa precipitation gauges revealed frequent 1 or 2 hour "gaps" in obviously prolonged precipitation periods, as well as frequent readings at the gauge resolution limit within the episodes. Examination of the Arizona data set revealed the same frequent 1 or 2 hour "gaps" in the precipitation patterns as in the Colorado data. At the Arizona gauge sites the threshold of detectability was .005 inch at all but the high resolution Happy Jack gauge where it was .002 inch. Figure 3.4 plots the cumulative distributions of episode durations for 217 episodes spanning 1466 hours and totaling 3-41 I i 'I I , 'I I I I , j I r I 'I t t i I