What does I30 tell us? An assessment using high-resolution rainfall event data from two Australian locations.

The maximum depth of rainfall in any 30-minute period - I 30 - has been used as an index of erosive rainfall events for more than half a century, and is a key component of the R factor in the Universal Soil Loss Equation (USLE) and its derivatives. Other indices are in common use, including I 5 , I 10 , and I 15. However, there has been little study of the ability of I 30 to capture the erosive nature of a rainfall event in regions beyond the foundation experimental work in the USA. This paper examines several apparently neglected issues: how reliance on fixed clock-periods might diminish the efficacy of the various indices of short-interval intensity; how well a single 30-minute period is able to reflect the intensity-duration properties of long enclosing rainfall events, and hence, whether the influence of event duration should be evaluated in applications of I 30 and related indices. Results from two Australian sites with high-resolution rainfall records show that within the enclosing rainfall events, actual intensities corresponding to the I 30 value (which is an equivalent intensity or 'rainfall rate', calculated from rainfall depth in 30 min) commonly occur for both less and >30 min. Often the I 30 interval within a rainfall event excludes the most intense periods of rain, and in long events the index reflects ≪1% of the event duration. The I 30 intensity can occur for total durations of 1–3 h in the events studied, or for just a few minutes. An alternative class of indices is proposed, in which fixed clock-periods are not used, and instead, the duration of rainfall within an event that exceeds a nominated intensity is recorded as an index of intensity. This has a number of advantages, including the ability to work with events shorter than 30 min, which are frequently intense but which cannot yield an I 30 index that is strictly comparable to those of longer events. Illustrative results for this new index are presented. • Rainfall events in two high-resolution rainfall records from two Australian sites are analysed for I30. • The I30 period sometimes includes intensity peaks but frequently does not. • Determining I30 for events shorter than 30 min in a consistent way is problematic. • A new intensity index, the duration of rain at and above the I30 intensity, is explored. [ABSTRACT FROM AUTHOR]