Rainfall Variability and Drought in the Hawaiian Islands

Video Transcript
Download Video
Right-click and save to download

Detailed Description

This webinar was conducted as part of the Climate Change Science and Management Webinar Series, held in partnership by the USGS National Climate Change and Wildilfe Science Center and the USFWS National Conservation Training Center. Webinar Description: Drought is a prominent feature of the climate of Hawai‘i with severe impacts in multiple sectors. Over the last century, Hawai‘i has experienced downward trends in rainfall and stream baseflow, an upward trend in the number of consecutive dry days and wildfire incidents, and regional projections show that unusually severe dry seasons will become increasingly common on the leeward sides of all Hawaiian Islands. This talk presents the state of the science on drought in Hawai‘i, and analyzes a new gridded drought index product (SPI, Standardized Precipitation Index) that was developed for the main Hawaiian Islands from 1920 to 2012. This dataset is used to determine the relationships between drought and large-scale modes of natural climate variability (including the El Niño-Southern Oscillation), analyze the spatial extent, frequency, severity, and duration of historical drought events, and examine trends through time. This spatially explicit analysis provides the historical context needed to understand future projections, and contributes to more effective policy and management of natural, cultural, hydrological and agricultural resources.


Image Dimensions: 1280 x 720

Date Taken:

Length: 00:43:04

Location Taken: HI, US

Video Credits

Abby Fraizer, USDA Forest Service Pacific Southwest Research Station - Institute of Pacific Islands Forestry
Elda Varela Minder and Holly Padgett, USGS National Climate Change and Wildlife Science Center


Emily Fort:  Welcome, everyone. We're so happy to have Abby Frazier here to talk about rainfall variability and drought in the Hawaiian Islands.

Dr. Abby Frazier is currently a post‑doctoral research geographer with the USDA Forest Service, Pacific Southwest Research Station and the Institute of Pacific Islands Forestry in Honolulu, Hawai'i, leading a new effort to understand the future of drought impacts in Hawai'i.

Dr. Frazier received her PhD and master's degree in geography from the University of Hawai’i at Mānoa, studying rainfall variability in the Hawaiian Islands, and has a BS in mathematics and a BA in geography from the University of Vermont.

Her research interests include climatology, geospatial analysis, geostatistics, climate variability, big data analysis, and landscape ecology. Thank you and take it away, Abby.

Abby Frazier:  Thank you so much for that great introduction and thank you all for tuning in this morning. I'm really excited to be here.

Before I jump into the content, I want to make sure to thank all of the folks who've made it possible for me to be here, in particular Dave Helweg and the Pacific Islands Climate Science Center for funding my post‑doc and parts of my PhD, Christian Giardina at the USDA Forest Service, USDA Climate Hub, and Tom Giambelluca at the University of Hawai’i at Mānoa Department of Geography.

Today, I'm going to cover a bit of background on climate in Hawai’i and rainfall in particular, and go over the motivations and objectives for this project. The rest of the talk will focus on two main topics, how different types of drought are expressed in Hawai’i and the results of a new gridded drought analyses that we have been working on.

Let me take you all to Hawai’i at this time. We are located in the Central Pacific and are made up of eight major islands. Oahu is our most populous island, with Honolulu labeled here. In the southeast corner, you see Hawai’i Island or the Big Island, where we have Hawai’i Volcanoes National Park and the famous Kona Coast, which you may be familiar with if you are a coffee drinker.

Our islands have extremely complex topography. If we look at the peak elevation for each island, you can see that we go from sea level up to over 4,000 meters or almost 14,000 feet. This complex topography, as you can imagine, leads to quite varied climate patterns.

Hawai’i is home to incredibly diverse climate patterns over a small area. We have moderate temperature and humidity, persistent trade winds coming from the Northeast. Thunderstorms are fairly rare, and we have extremely dry high elevation areas, above 7,200 feet.

These diverse climates lead to a wide range in biomes. We have tropical rainforest, all the way to barren alpine zones.

None of these climate variables exemplifies our diverse patterns more than rainfall. This is a map showing mean annual rainfall for the state of Hawaii. When we put in our prevailing winds, our trade winds, coming from the Northeast, these interact with our mountain slopes, producing wet windward areas and dry leeward areas.

Most places in the state have a wet season from November to April and a dry season from May to October.

If we look at the range in rainfall values here, we have a greater range in rainfall in the state of Hawai’i than you find on some continents. We have places in the rain shadow areas that get less than 10 inches of rain per year.

In Honolulu, we receive about 25 inches of rain a year. Then in the mountain summits, we get places on Oahu that receive 250 inches of rain a year. North of Hilo on Hawai’i Island, about 300 inches of rain a year. In the center of Hawai’i, we get almost 400 inches of rain a year.

Our wettest place in the state is actually on East Maui, a site called Big Bog. This area receives 400 inches of rain per year on average, making it not only the wettest place in Hawai’i, but the wettest place in the United States.

When people think of Hawai’i, there's often this perception of lush, beautiful rainforest and wettest place in the United States, so how important is drought in a place like Hawai’i?

In 2010, we had our driest year on record, at least back to 1920, followed by 2012, which was our second driest year. According to the U.S. Drought Monitor, over 50 percent of our land area experienced moderate to extreme drought conditions from mid‑2008 to January 2014.

This was an extremely long and dry period. It had impacts across the state in many different types of sectors. This particular drought event, because of its multi‑year duration, opened a lot of people's eyes in the state to the fact that Hawai’i is actually fairly vulnerable to drought.

A lot of questions arose after this 2010 drought period. Is this 2010 drought a normal drought? How does it compare to previous drought events? Are droughts getting worse? Is this drought an example of how droughts will be in the future? How will drought conditions change in the future? What can we expect?

We realized the last real comprehensive drought analysis for Hawai’i was done in 1991. We now have a lot more data sets available to us, in particular, gridded climate data. A new drought analysis was needed to understand what the impacts have been in the past, to establish that baseline to prepare for future droughts.

The objectives of the projects for my postdoc are, first, to complete a literature review from all previous drought studies that have been done, to update that 1991 report. Secondly, to do a new retrospective analysis of meteorological drought conditions using a new drought index product, which I will get into.

Third objective is to use downscaled climate projections to determine the future of drought probabilities in the state.

Hawai’i has a very small spatial extent. We rely heavily on the use of downscaling from the global climate models because our whole state tends to fit in just a couple of grid cells in the global climate models, so they're too coarse to understand what's going on in the state.

Lastly, we want to better understand resource managers' decision‑making processes during drought and responses to drought. We're hoping, with this project, to conduct a survey of land managers in the state so that we can understand what types of data that they're using and how we can provide better data products for them.

For this talk, I'm going to focus on the first two objectives here.

To go through the previous literature, I'm going to talk about how drought is expressed, using the five different types of droughts that are typically recognized. Most of you have probably seen this.

We categorize drought often into meteorological drought, which is when you first get that precipitation deficiency, and you get this low infiltration and reduced runoff. This is often accompanied by low relative humidity, high temperatures, and increased evapotranspiration.

As that meteorological drought persists, you get soil water deficiency. This tends to lead to plant water stress and reduced biomass and yields.

Again, as time progresses, and that meteorological drought persists, this eventually impacts the R hydrology, so we get reduced stream flow, and, potentially, reduced groundwater recharge. This has impacts on our economic and social impacts, and impacts in our ecosystems.

I'm going to focus on the physical types of drought today.

Hawai’i has a very extensive network of rain gauges in the state. We're extremely fortunate to have this great data set. As of about 2010, we had around 430 gauges operating in the state, and you can see their spatial spread here on the map, but this number has changed over time.

It actually peaked in the late 1960s, when we had over 950 rain gauges operating at the same time in the state, which is an incredible number of gauges. This is mostly due to the plantation agriculture.

As the plantations, the sugarcane and pineapple, expanded, they installed more rain gauges. As they began to close down, they discontinued many of those gauges. This expansive network of data allows us to monitor the meteorological conditions in the state.

We can't talk about drought and rainfall without considering natural variability. Our climate in Hawai’i is strongly influenced by large‑scale modes of natural climate variability, such as the El Niño‑Southern Oscillation, the Pacific Decadal Oscillation, and the Pacific North American.

ENSO, in particular, due to our location in the tropics, and the fact that the phases recur pretty frequently every three‑to‑seven years, has very strong seasonal impacts on our rainfall.

If we look at the seasonal rainfall anomalies for the state of Hawai’i during strong El Niño events on the left, and strong La Niña events on the right, we can see that in winter months, December, January, February, El Niño events tend to give us anywhere from 10‑to‑80 percent below normal rainfall, while La Niña events give us 10‑to‑50 percent above normal rainfall. You see the pattern is opposite in the summer and fall months.

With a strong El Niño, in the summer and fall, we typically have increased tropical cyclone activity near the Hawaiian Islands. This brings us wetter summer and fall conditions as we head into a dry winter, and for La Niña it's the opposite. We tend to get a dry summer, followed by a wet winter.

In Hawai’i, most places have their wet season in the winter, and so drought events are typically associated with El Niño events. We expect when we're getting an El Niño event that we're going to have a very dry winter.

We also have to consider all of this variability within the context of baseline change, and in Hawai’i there's evidence for rainfall trends decreasing.

This is a 500‑year reconstruction of winter rainfall for the past 500 years, and we see a drying trend in the last about 160 years. When we look at the spatial patterns of that drying, the spatial patterns are heterogeneous.

This is looking at a rainfall trend map for the dry season from 1920 to 2012, and the red colors are showing you decreasing trends, the blues are increasing trends, and anything in the black hatching is statistically significant. We can see, especially on the leeward side of the Big Island, a very strong decreasing trends in the dry season, but the patterns like I said are not homogeneous.

For future rainfall, the projections right now are showing that the leeward areas are expected to get much dryer, and you can see this here in the statistical downscaling annual rainfall projections. The windward areas are expected to remain the same or see slightly wetter conditions.

There's a little bit of discrepancy between the different methods, but in general we have a consensus that leeward areas are expected to get much dryer, and these were already highly drought prone regions.

Another trend we've been seeing is in consecutive dry days, which is sort of a metric for drought, and a study from 2010 showed that most stations around the state have experienced upward trends in their annual maximum CDD or consecutive dry days over this period.

A more recent study has shown that the leeward areas in the state showed the largest increases in consecutive dry days. We're seeing dryer trends across the state, but in particularly, leeward areas are highly vulnerable.

In terms of agricultural impact of drought in Hawai’i, farmers and cattle ranchers are some of the community's most immediately affected by drought. State and county governments pose voluntary and mandatory water restrictions, and typically for irrigation there are mandatory water restrictions, and non‑irrigated areas tend to receive lower crop yields, reduced ground cover and pasture.

Farmers often have to import water. They see crop loses, lose significant revenue, and often take out loans.

There was a study in 2010 that interviewed folks in the agricultural sector, and found that many farmers find that it's difficult to make decisions based on experience when current conditions no longer seem to reflect the past.

It's becoming increasingly difficult for farmers in the state when we have these dryer conditions. Cattle ranchers often see increased cattle mortality and reduced calving rates. They have to reduce herds, leave pasturelands, and they experience economic losses for several years following drought.

In 1999, in the cattle industry alone, the revenue losses were around $6.4 million and the following year another $9 million in revenue losses, so drought has a major impact for the farming and ranching communities here in Hawaii.

It also has large impacts on our water supply. 99 percent of our drinking water comes from ground water in the state of Hawai’i, and the most important aquifers consist of a fresh water lens that floats on more dense seawater.

For a thin aquifer, like in the Kona region, these tend to be more vulnerable to drought and to periods of low rainfall, because the transition between the fresh water and the salt water is shorter, and you have higher likelihood for salt‑water intrusion. In the thicker aquifers, like the Pearl Harbor aquifer on Oahu, these are less sensitive to periods of low rainfall.

However, when you have increased water demand due to drought, you end up with higher pumpage and the water level in these thick aquifers can decline.

We've also seen, in the stream base flow across the state, long term declines over the past century or so, and these are likely related to a decrease in ground water storage and recharge. This serves as an indicator of decreasing water availability across the state, so likely related to the decreasing rainfall trends.

Ecological impacts in Hawai’i ‑‑ we actually don't have a lot of documented impacts in the literature.

A lot of the examples that I'm about to show are individual examples that we've heard from resource managers across the state, but there's not a lot that's actually published in peer reviewed articles here, but we have evidence for ungulate problems being exacerbated during droughts.

The feral ungulates change their foraging patterns and encroach onto pasturelands and into populated areas. We also see pest problems exacerbated, and this in some cases can lead to forest loss and mortality of native trees, but we don't know the extent of how much drought‑induced tree mortality we can expect with future drought. We need more studies on that.

Drought also threatens rare and endangered species. It leads to invasion by weeds and other non‑native species, and the decreased top soil often leads to increased erosion and sediment delivery to the nearshore areas, so this is also impacting our coral reefs in the nearshore areas, and we see increases in wildfire.

This is probably the most dominant ecological impact of drought in Hawai’i, and wildfire is a major threat to Natural Resources and native species in the state. Over the past century, we've seen an increasing trend in annual area burned. The largest areas burned are often in dry non‑native grasslands and shrublands.

When we compare the percent area burned in Hawai’i to the U.S. mainland and just the Western States in the mainland, over this period of 2005 to 2011 Hawai’i actually had a greater percent area burned than the mainland. Wildfire is a major threat here in Hawai’i and wildfire is linked to El Niño events.

That wetter summer and fall that I mentioned previously, increases fuel loads, especially in dry areas across the state. Then you head into the dry winter, which increases the potential for wildfire occurrence and spread.

When we know that we're supposed to experience an El Niño event, the folks, especially at Pacific Fire Exchange, do a great job of doing lots of outreach and education to alert folks to this pattern and try to reduce human ignitions.

There's also some evidence from remote sensing studies about how vegetation in Hawai’i responds to drought. There was a statewide study looking at wet and dry forests during an El Niño‑induced drought. They found that the dry forests browned‑down during the drought, while the wet forests greened up.

Their theory for why this was happening was because these wet forests are typically light‑limited. They're cloud forest areas, that when we get a drought and the clouds are no longer there, they can actually photosynthesize more, so we see this greening signal in the wet forest.

Another recent study that looked at an area on windward Maui looked at four different drought events and compared two different remote sensing products in a wet forest and a shrubland, and found that the first three droughts that they looked at showed an overall greening signal consistent with the statewide study that during drought, the wet forests are light‑limited and then they green up.

During this last drought period, the 2008 to 2014 period, both the wet forest and the shrublands showed browning and so likely the long duration that multi‑year nature of that drought overrode the light limitation and water became the limiting factor.

We also have new evidence showing that on Hawai’i Island, long term drying trends, so that long term baseline change, these trends are driving a large‑scale decline in forest canopy volume and greenness. This long‑term drying is affecting our forest structure and functioning.

We can see the red here in the leeward parts of the Big Island, looking at mean photosynthetic activity. There's evidence not only that short‑term drought events are causing browning and stress in our ecosystems, but also that this long‑term drying is having impacts.

For the rest of the talk, I want to focus on this new retrospective analysis that we've been working on.

To analyze drought in Hawai’i, we need to find a good source of data. The U.S. Drought Monitor is the most widely used product for current and past drought conditions in the state.

It's updated every week; it's posted online; it has these really well‑defined drought categories that are easy to interpret, but the spatial detail isn't great.

It's fairly smooth and the product only goes back to 2000. If you're trying to analyze drought prior to the year 2000, this product isn't going to work very well.

We also have KBDI that's calculated here in Hawai’i. This is a really useful metric for fire managers, but right now it's only calculated at one station at the Honolulu Airport. This is not very representative of statewide drought and fire conditions.

In terms of data that you would need to calculate PDSI or SPEI, these more complex metrics of drought, we don't have enough data on soil moisture or evapotranspiration in the state to be able to calculate these.

What we do have is a lot of great rainfall data, however. We can use an index like the SPI, the Standardized Precipitation Index. This is one of the most commonly used indices around the world for looking at drought, and it's great because it allows you to look at drought over multiple time periods. You can look at a one‑month, a three‑month, a six‑month SPI.

For example, a three‑month SPI, in August of 1990, would look at the June/July/August 1990 rainfall and compare it to all June/July/August rainfalls in every year, so your sample needs to have at least 30 years.

It's standardized and normalized so you can look at wet and dry periods. The units are just number of standard deviations or Z‑scores. Any positive values are wetter than the mean and the negative values are drier than the mean, so you can pick out drought events when it dips below zero for a long enough period of time.

It has very nicely defined definitions of what constitutes a mild drought, a moderate drought, a severe drought, or an extreme drought. This is a very appealing index. What we've done in Hawai’i is using monthly rainfall maps from 1920 to 2012.

Matt Lucas, who's a student at the University of Hawai’i here, calculated the SPI at every 250‑meter pixel in the state and then mapped the SPI values. He did this for 10 SPI time periods, from 1 month up to 60 months. This now allows us to calculate the frequency, variability and trends in drought back to 1920 and to identify drought events through space and time.

When we look at this 250 meter spatial resolution compared to the Drought Monitor, we get much better spatial detail. This is a look at the six‑month SPI maps for the Big Island from January 2008 to December 2012. Remember, this was that extremely dry period that I mentioned earlier. You can see where the drought grows and which areas it starts in.

Around the summer of 2008 is when we start to see some strong drying going on, and then it spreads and the entire island is experiencing drought. Then we get some relief in 2009 but the El Niño event we get in 2009‑2010 leads to some pretty strong drying especially on the leeward side.

By 2010, you can see we have extremely dry conditions across the Big Island. Again, it subsides a little in 2011 and again in 2012, we get hit with more very, very dry conditions.

Some of the preliminary results that we've calculated are mean SPI by decade. This allows us to identify wet and dry decades and periods overall and the spatial patterns there. One thing we've picked out from this is that the last decade has been the driest on record, with statewide drought conditions present 90 percent of the time between 2006 and 2012.

When we look at proportion of time in drought, or sort of a frequency metric, from zero percent of the time in drought to 100 percent of the time in drought, we see that the last decade, again, you have places that are in drought about 80 percent of the time during this decade.

We can look at metrics like that. We can also look at the number of severe and extreme drought months by decades, so the number of months where the SPI was less than negative 1.5, and we can see pretty low drought statistics for the first half of the century.

Then starting in the 1970s, we see a real ramping up of number of dry months that we have statewide. If you pay attention to this last decade column, 2010 here only contains three years of data and has a higher number of severe and extreme drought months than we've seen in the entire rest of the records.

We can look at that same statistic by Island to get a feel for some of the spatial patterns and we can see that the 2010 decade isn't the driest necessarily for everyone, but the Big Island was especially hit hardest by the dry years, in 2008 to 2012. We can see the spatial patterns here a little bit.

What we're working on right now is identifying events, so all historical events, and getting a sense for the spatial extent of each event, the duration, the frequency, and severity and hopefully getting maps like this and better seeing where the drought started and how it grew.

This is going to allow us to answer questions, such as what was our most severe event? What was our longest event? Are there differences or trends by island or through time? What is the role of El Niño and La Niña? How are the droughts related to El Niño and where are the most drought‑prone areas? Where are the highest drought frequencies?

We'll be able to answer all of these questions once we can pull out all of these events, and be able to place that 2010 drought into context. We also, with this geospatial product, can provide customized drought data for different conservation lands in the state.

We're planning to extract the drought data for national parks, for natural area reserves, for private lands, DoD lands, pretty much anyone who has a land area that they want to look at. We can extract the SPI data and provide a customized drought history for that area to help with management efforts.

I'll go through an example where we've actually done this so far. This example is based on windward Maui and we call the study area the forest line. It's the upper limit of the cloud forest where it transitions into sub‑alpine shrublands.

Again, it's on windward Maui and it corresponds with the upper limit of the bird habitat. Why this study area is particularly interesting is because we have not only contemporary evidence but also paleoecological evidence that El Niño induced drought events drive forest line position.

We extracted the SPI data for the forest line on Maui and we identified 28 drought events from 1920 up to present. For each event, we calculated the intensity, the magnitude and duration. The magnitude is plotted here for the 28 events that we've found.

The last drought, this 2008 to 2014 drought had the longest duration. Average duration is about 15 months. This drought had a duration of 70 months. It has the highest magnitude and highest peak intensity of any other drought in the record.

When we lined up these 28 droughts from the forest line with ENSO events, so strong, moderate, and weak El Niño events and strong, moderate, and weak La Niña events, we tried sorting based on the different metrics to try to understand the relationship.

In general, we found that of these 28 droughts, about two‑thirds of them were associated with El Niño events, but about half were associated with La Niña events, which I don't think we were quite expecting.

When we sort the 28 droughts based on their intensity, we see that some of the most intense droughts were associated with weak El Niño events, while some of the least intense droughts were associated with strong El Niño events.

There wasn't really a clear pattern based on drought intensity. When we sort based on drought duration, again, we see some of the shortest duration droughts are associated with strong El Niño events. One of the longest droughts was associated with a moderate La Niña.

This kind of goes against our idea that a strong El Niño event always leads to a strong drought. One thing we did notice was that the two longest droughts in the record, that 2008 to 2014 and the 1971 drought, were associated with three ENSO events, back to back to back ‑‑ some kind of La Niña, followed by an El Niño event, followed by a strong La Niña.

When we looked into the actual rainfall during those events, we found something interesting. This map is showing the average rainfall anomaly for all strong La Niña events in our record. This, on average, is what strong La Niña events do to our rainfall.

You can see our forest line study area here. Typically, we get wetter than average winter conditions during La Niña. If we look at the two strong La Niña events associated with those two longest drought events, we can see that in the forest line area on Maui that these two events did not lead to the typical rainfall response, that during these strong La Niñas, we didn't get the wetter‑than‑average conditions that we were expecting.

This leads us to the conclusion that we need to better understand the stability of the ENSO rainfall relationship, and how has this been changing. There has been a recent paper showing that since the 1950s, rainfall during La Niña events has been decreasing.

The other thing that this tells us is that the sequence of ENSO events matters. If you get an El Nino event, so you have a dry winter. You then follow that with a La Niña, so you get a dry summer. Then instead of getting a wet winter, you get another dry winter with this type of La Niña that we saw.

Then you're going to have dry, followed by dry, followed by dry. The sequence of these ENSO events could be driving these long duration droughts.

With that, I'll wrap up. Hopefully, I've convinced you that drought is a prominent feature of Hawai'i's climate, with severe impacts in multiple sectors, and that we've been seeing long‑term drying in the state that's expected to continue in these leeward drought‑prone areas.

This new analysis that we're working on will update this 1991 report and identify data gaps. We're hoping to provide high‑resolution drought information, identify the spatial details of ENSO's influence on drought, improve our understanding of ecological drought, look at drought changes in the future, and conduct a manager survey so we can provide more relevant data products.

With that, I will say thank you and I'm happy to take any questions.

John Ossanna:  Thank you, Abby. I see a few people...Richard, I see right there. Richard, if you're on the phone, go ahead and press star six and ask your question.

Richard:  Can you hear me?

John:  Yes.

Richard:  Great talk, Abby. I did have a question about why you chose the specific SPI time scale that you did. Why did you chose the six‑month instead of another time scale, say, the three‑month or the nine‑month.

Abby:  This was just an example of one that we picked out. We're planning to do the multiple different time scales and see what best correlates with what's happening on the ground.

Six months gives us an idea of the seasonal changes that are going on and starts to give us an indication of when we're going to see hydrological impacts. We're planning to pull these out for different SPI time periods as well.

Richard:  Thank you.

Abby:  Thanks.

John:  It looks like Paul put in a question. It appears that trade winds are a key factor in rainfall pattern. Has there been analysis of trade wind patterns and how they are affected by climatic drivers?

Abby:  There have, actually, thanks Paul. There was a recent study that showed that our trade winds have actually been shifting in direction. Instead of being predominantly northeast, they're shifting to a bit more east‑northeast.

This is part of a regular cyclical pattern. Every 20 years or so, our trade winds tend to do this sort of shift. I don't think anyone's taken it the next step further to really analyze how these shifts in the trade winds are actually possibly driving some of these changes in rainfall or how that's related to ENSO.

There've been some studies of the data on the ground, but I don't think anyone has taken it to the next step yet.

John:  We have a question from Jonathan. The Hawai’i City Council tomorrow is considering a water use and development plan update for the Keauhou aquifer. I hope I said that right. I probably didn't. It is the main planning document for water use for the state and county in the area.

The report does not issue or use the word drought, not talk about the concept. Do you have any reaction?

Abby:  Wow. That's a little bit surprising, especially considering that West Hawai’i was one of the hardest hit areas in this last drought or in the 2010 drought that we had. Again, in 2015, 2016, we had another drought. Again, West Hawai’i was hit pretty hard by this. I find it surprising.

Hopefully they're at least, if not considering drought events, hopefully, considering the long‑term drying that has been shown in that area, because I think maybe more so than drought, this long‑term drying would have big impacts on that aquifer.

John:  I see a few people are typing away.

We have a question from Nāmaka. Are you considering the impacts of the current eruption of Kilauea and the associated vog on drought conditions in West Hawaii?

Abby:  Hi Nāmaka, thank you so much for that question. That's actually one of our leading hypothesis for why we've seen such strong declines in rainfall in West Hawai’i is due to the eruption of the Kilauea Volcano.

It's been erupting pretty consistently since 1983. A lot of that vog, or volcanic smog, gets carried over to the leeward side of the islands. What it does is it acts as a cloud condensation nuclei and spreads the cloud droplets out too thin. They're not heavy enough to actually fall as rain.

This is one of our leading reasons for why the rainfall has been declining so steadily in this region.

John:  We got tons of questions today. This is great.

Has there been an analysis of the quality of data used for the maps that drive the SPI analysis? Hundreds of rainfall sites, the quality must vary between stations.

Abby:  Yes, that is absolutely true. The quality does vary between stations. As part of our lab in the Geography Department, here at UH Mānoa, basically all of my grad school career has been doing data quality control on these stations across the state.

I think, in total, we had about 2,000 gauges that have operated at some point in our network. We have done a lot of work quality controlling the data, homogenizing the data, looking for any big spikes and things. The rainfall maps that were used in the SPI were actually created by me. [laughs]

There was a lot of work that went into deciding on the interpolation method to use and quality controlling these maps as well. We've done our best to reduce any errors due to data quality.

John:  We have a question from Martha. Did you look into what type of rain events were missing during drought periods? Was it the larger winter storms or were trade showers weaker?

Abby:  That's a really great question. It's something that I would like to look into in the near future. A couple of other folks at the university have been looking at storm events in Hawai’i and how these have been changing through time.

What they've been finding is that atmospheric conditions in the Pacific have been less favorable for big storms that come through, like Kona low events, which are low‑pressure systems that just sit over the islands and dump rain.

We've been seeing fewer and fewer Kona low storms over recent years. There's also some evidence that as temperatures rise, that mid‑latitude storm systems are going to move poleward.

We already catch the tail ends of those storms. If they move further north from us, we will miss out on more of those storm rains, but I don't think we've been able to detect any differences in trade showers yet. That would definitely be something to look into, to better understand what's driving those.

John:  Thank you, Abby. It doesn't appear that we have any further questions. I just want to say thank you again for a great presentation. Everyone that was on the line, thank you for your participation. Be on the lookout for our next webinar in this series.

Abby:  Thank you so much, everyone. Feel free to email me with any further follow‑ups.