Mauka Showers

We’re nearly at the end of September, and with it will be the end of the May through September 2025 dry season (kau) in the Hawaiian Islands. Ancient Hawaiians actually considered the dry season to be from May through October. However, many meteorologists and climatologists now classify the dry season to be from May through September, and the wet season (hoʻoilo) to be from October through April (Sanderson 1993, Blumenstock and Price 1967). I’ve shown in previous posts a graph of monthly average rainfall at Mānā, Kauaʻi (below). As a leeward site, it captures the change from dry season to wet season with a sharp increase in rainfall from September to October. Of course, one could also argue that maybe the wet season should be considered to be done in March, but I’ve never seen that come up in any documentation on when the wet season ends. The October start to the wet season also makes sense because mid-October is when the first cold front (“cold” being used loosely) reaches the main Hawaiian Islands on average.

Graph of monthly average rainfall in inches at the Mānā gage in west Kauaʻi. The data used in the graph are from the Rainfall Atlas of Hawaiʻi.

The higher rainfall within the wet season is due to the effects of stronger atmospheric disturbances that move over or near the state during this time of the year. These disturbances can come in several forms, such as kona lows, upper tropospheric lows, or the above mentioned cold fronts. Note that “kona” in kona lows is used as an adjective (meaning “leeward”), and is not capitalized. It doesn’t refer to the Kona area of the Big Island. This follows the convention in Simpson’s 1952 seminal paper on the topic.

Some of this rainfall can be intense enough to produce flooding. The National Weather Service (NWS) classifies excess freshwater to be either floods or flash floods. The term flash flooding is used for situations where flooding occurs within six hours of the cause (usually intense rainfall). In the Hawaiian Islands, impacts occur well within the six hour time frame so all floods in the island chain are considered to be flash floods. The rapid response time is mainly due to the small size of the drainage basins in the state, as well as the high intensity of the rainfall.

Flash flooding events can occur during any month of the year, as seen in the graph below. The data displayed here are not from official NWS records, but are from my own event log that I’ve kept for many years to more easily keep track of trends and impacts. If I wanted the official NWS numbers, I would have to request the data from the Honolulu Forecast Office. Their Warning Coordination Meteorologist (WCM) would then have to login to the NWS Performance Management site (accessible only to NOAA staff), pull the events from the database, and send the bulk file to me. I would then have to parse it out and do some additional processing. Since the WCM has better things to do, and because my data will likely be pretty similar to the official numbers, why go through that hassle just to have the “official” stamp on it when I’m just using the numbers in a blog post? Events in my log are defined by impacts, such as, but not limited to, public road closures, streams overflowing bridges, or flooding that results in property damage.

The most active month is clearly March, followed by December. June is easily the least active for flash flooding. This, combined with the low frequency of tropical cyclones make June the best month in terms of low probabilities for significant weather events. The graph hasn’t always looked like this. In a version covering the period from 1960 to 2000, November was the clear #1, followed by October. However, over the past decade or so, October and November have had fewer impactful flash flood events, allowing other months to catch up and surpass their totals.

Number of flash flood events by month for the period from 2004 through 2024 across the Hawaiian Islands.

As you probably know, rainfall is highly variable from year-to-year, so the annual number of flash flood events also has a high degree of variability. The graph below shows the number of events across the state by year. There does not appear to be an obvious trend in the annual totals through the 21-year period. Of course, you could read this after a nice Happy Hour at the bar and see all the trends you want, but none are obvious to me right now. There are two obvious outliers, 2006 and 2018, with 2018 having a slightly higher number of events. Both years started with ongoing weak La Niña events, then transitioned into weak El Niño events. However, they had a very different distribution of events throughout the year (2nd graph below).

In 2006, the flash flood events were mainly concentrated within a very wet period from late February through early April in what is sometimes referred to as the “40-Days of Rain”. Personally, I never liked the term because it didn’t rain significantly every day. In fact, there were some days within that period that had really beautiful weather! But the term was somehow let loose into the wild and propagated. In recent years it seems to have been mostly forgotten, but every once in a while I hear, “oh yeah, you remember the 40-Days of Rain? It rained hard for 40 days straight!” and I have to try hard to not roll my eyes. In 2018, the monthly totals were more evenly distributed through the year but had some highly noteworthy events. These include the April event at Waipā, Kauaʻi that broke the U.S. record for the highest 24-hour rainfall, Hurricane Lane in August that produced the second highest tropical cyclone rainfall total in the U.S., and the landfall of Tropical Storm Olivia over Maui County in September.

Number of flash flood events by year for the period from 2004 through 2024 across the Hawaiian Islands.

 Number of flash flood events by month for 2006 and 2018 across the Hawaiian Islands.

The last graph I wanted to show breaks down the number of events by county. The totals do not appear to be a function of the land area in the county or the average rainfall. Since my numbers are based on impacts, they are more related to flood vulnerability. For example, Kauaʻi has a total of 110 flash flood events. Of these, 82, or about 3 out of 4 events, involve the closure of Kūhiō Highway due to the overflow of the adjacent Hanalei River either as the sole flooding impact or among several impacts across Kauaʻi County. I’ve included an image of a recent Hanalei River flash flood below to 1) show you what a typical Hanalei River impact looks like, and 2) give you a break from the incessant graphs in this week’s and last week’s blog posts. In the case of the City and County of Honolulu, which has the highest tally of events, it is more a matter of having a lot more things to flood. The dense population on Oʻahu means there is a lot more infrastructure that can be impacted. Kamehameha Highway at Waikāne Stream is the location that is flooded most often, and is involved in roughly one out of three of the 111 events in my log. The third image below shows Kamehameha Highway covered with water from Waikāne Stream, but not deep enough to close the road.

Image from the County of Kauaʻi showing Kūhiō Highway inundated by the overflow of Hanalei River on July 20, 2025.

Kamehameha Highway inundated by the overflow of Waikāne Stream.

Now that I’ve set the stage with the flash flood numbers, over the next few weeks I’ll cover the products issued by the NWS during heavy rain and flash flood events. Hopefully, these posts will provide some useful information prior to the bulk of the heavy rain events in the upcoming wet season.

References

Blumenstock, D. L., and S. Price, 1967: Climates of the United States – Hawaiʻi. Climatography of the United States No. 60-51, U.S. Dept. of Commerce, ESSA, 27 pp.

Sanderson, M. (ed.), 1993. Prevailing Trade Winds. University of Hawaiʻi Press, Honolulu, Hawaiʻi.

Simpson, R. H.,1952: Evolution of the kona storm: A subtropical cyclone. J. Meteor., 9, 24-35.