Antecedent soil moisture conditions and receiving rainfall predict high streamflow in flashy watershed in Hawaiʻi

Abstract

Extreme events such as heavy precipitations and associated floods often have devastating consequences on societies and ecosystems. However, extreme rainfall alone is not the sole driver that results in high flow. Previous studies highlighted that annual maximum daily rainfall exhibits inconsistency with annual peak discharge in their occurrence timing in Hawaiʻi. The mechanism of runoff generation and, therefore, consequential storms and floods remain unclear. In this study, we investigated the linkage between extreme rainfall and high discharge events. Rainfall, soil moisture, and discharge data, in one watershed on Oahu and one on Maui were used in this study. We defined antecedent soil moisture conditions using Antecedent Soil Moisture Indexes (ASI) calculated from soil moisture data. We compared the timing of the occurrence of annual maximum hourly or accumulated (from three to twelve hours) rainfall and annual peak discharge. Then, we estimated the timing of high-flow events based on antecedent soil moisture conditions and maximum hourly rainfall. Multi-linear regressions were used to estimate high-flow event timing. Finally, we compared these estimates with the actual high-discharge events. We found out that the consistency between the timing of maximum rainfall and the timing of annual peak flow did not improve when we used hourly or accumulated hourly rainfall. Nevertheless, the consistency improved when we included antecedent soil moisture conditions by including ASI. We successfully estimated the occurrence timing of majority high-flow events at the site on Oahu and at the site on Maui. These accurate estimations emphasize the importance of incorporating soil moisture with hourly rainfall to estimate high discharge events and increase our understanding of flood events induced by extreme rainfall in Hawaiʻi.