An El Nino winter for coastal Washington

El Nino intensity forecast from here. 

I've been getting a lot of questions about this winter's El Nino forecast, and in particular what it may mean for coastal flooding and erosion in Washington State. I don't have fantastic answers to that question, but if a strong El Nino does develop (generally associated with the Oceanic Nino Index exceeding 2 degrees C), and past is prologue, it probably does mean something for those hazards in Washington. Those questions also sent me down a bit of a rabbit hole focused on coastal water level here in Washington in particular, and specifically what is referred to as the "non-tidal residual", or NTR, which is the difference between the astronomical tide, and the water level actually measured on the coast by tide gauges.  

The non-tidal residual, or NTR, measured at the tide gauge in La Push

For us in Washington, the NTR is due generally to weather - coastal water level tends to be higher during storms, and lower, and even negative, during nice weather with high pressure.  What this means on the ground is that the actual tide is a higher than predicted by your tide chart when the NTR is positive, and a lower than predicted when the NTR is negative. The NTR as measured at our tide gauges is therefore a useful metric for generally evaluating the potential impacts of an El Nino on coastal flooding and coastal erosion.  

Winter (Nov-March) NTR distributions from Seattle (top panel) and the ONI index for the same winter months in the bottom panel. The dashed line in the bottom panel is the 2 degree Celsius threshold that generally is used to characterize a "strong" El Nino

I grabbed a few datasets I had previously compiled of hourly water level from some of our tide gauges here in Washington, specifically La Push, Cherry Point, Seattle and Port Angeles, and also the ONI index compiled by NOAA to just quickly evaluate the relationship between the ONI and non-tidal residuals for those gauges. I started by generating a summary figure like the one shown above, to sort of visually see the distributions of both datasets for winter months (I chose November - March). I then plotted the relationship for each gauge for the mean of both of those datasets for each winter in the record. I will focus on those.

Average winter NTRs from Seattle versus the average ONI index

Generally, the story from all of the locations is similar.  Three strong El Ninos are in the ONI index record: the winter of 1982-1983, 1997-1998 and 2015-2016.  All three of these winters cluster at the top of these plots with the largest mean NTRs on record, hovering between 15 - 20 cm. In practical terms: On average during those winters, the actual measured coastal water level in Washington was 6-8 inches above what is predicted in the tide chart. That doesn't seem like a lot, but a few additional inches of average coastal water level can have a considerable influence on the likelihood of flooding.  

Average winter NTRs from Port Angeles versus the ONI index

But there are some nuances that make it hard to know exactly what is coming this winter.  There isn't a perfect relationship between the ONI index and the mean NTR...the 2016-2017 El Nino winter was the strongest on record based on the ONI index but has the lowest mean NTR of the three we can see in our tide gauges. The opposite is true of the 1982-1983 El Nino, which has a lower ONI index but the highest NTR. There is also spatial variability - the NTRs in Port Angeles associated with the 1982-1983 El Nino and the 1997-1998 event are a bit higher than those in Seattle. Some of this might be explainable, and it is worth noting here that I haven't removed sea level rise from the tide gauge data I've used here, and also didn't use the new Relative ONI, but I think the point still stands...for us here in Washington, a strong El Nino is likely to increase our odds for higher than average coastal water levels, and the coastal flooding and erosion associated with it, but it is hard to know exactly what to expect. Furthermore, El Ninos influences far more than just coastal water level - I'm focused here on just a tiny slice of the potential impact "pie".

Has there been more coastal flooding on the Strait of Juan de Fuca recently?

 

Three Crabs Road on the Dungeness River Delta on 1/3/2026.  Photo by Rob Casey 

I feel like I've been having numerous conversations in the past few months regarding whether coastal flooding in the low-lying zones on the Strait of Juan de Fuca, particularly the neighborhoods along the Dungeness River Delta, have been occurring more frequently in recent years.  

And the answer, after at least a preliminary look at some tide gauge data, seems to be yes. 

Here is how I approached this. First, there isn't a tide gauge or other water level measurements from the Dungeness River delta, at least that I have access to, but the tide gauge at Port Angeles seems to do a pretty good job in terms of reflecting what happens at the Dungeness River delta. During a 2018 flood event, for example, I was able to relate the surveyed elevations of flood lines along Three Crabs Road to water levels measured in Port Angeles. 

Another one from Rob Casey from 3 January 2026, taken from here looking north.

My analysis of the 2018 event also provides some insight about how high the water has to get to cause flooding on the Dungeness River delta. But I actually think that the photo at the top of this post, from here during an event on 3 January 2026, is also useful for that purpose.  First, the coastal water level on 3 January 2026, when this photo was collected, reached 2.72 feet relative to MHHW, just high enough to send water across the road. Next, it also generally corroborates the flood elevations from the 2018 event I wrote about here in my blog. In my mind, therefore, and for the purposes of what follows, I viewed that coastal water level elevation of 2.72 feet MHHW as a useful threshold for evaluating the frequency of flooding along the Dungeness River delta - when water measured at the Port Angeles tide gauge reaches or exceeds 2.72 feet MHHW we can assume that flooding is likely happening at least along parts of the Dungeness River Delta.

Therefore, to get a sense for how often water makes it high enough on the Dungeness River delta to start flooding things like roads, and whether that has occurred more frequently recently, we can ask a relatively straightforward question: How often has the water measured at the Port Angeles tide gauge exceeded that selected threshold of 2.72 feet relative to MHHW? 

Here is the entire coastal water level record from Port Angeles, dating back to 1979, where the water level is shown as feet relative to Mean Higher High Water (MHHW).  I've added a red dotted line at that key 2.72 feet MHHW threshold:

and here is a zoom into just the top of that figure, with a few annotations, to hopefully highlight when those extremes occurred:

On this plot there are two distinct time periods that I've marked with arrows, the winters of 1982-1983 and the winters of 2003-2006, during which there were a number of water levels measured at the tide gauge that were high enough to presumably drive flooding on the Dungeness River delta. There was presumably some flooding during those two time periods, but I am unfamiliar with reliable records documenting flooding during those time frames. The winter of 2006, at least, was notable for driving significant storm damage at Seashore Lane, which is documented in the Washington's Marine Shoreline Design Guidelines. 

After that 2003-2006 time period there is roughly a decade of relative quiet, with just one event in November 2011 that exceeded the threshold that I'm using here. 

But the main takeaway...from 2015 to the present though there are five events that exceed the threshold, one every other winter on average. That seems to square with the general feeling the people seem to have that there have been more flooding events than usual in that area. Four of those events fall on to this list on NOAA's coastal inundation dashboard for the Port Angeles tide gauge, which is intended to capture the top 10 events on record for the tide gauge. 

This begs the question...why?  I don't have a great answer to that, but there are two candidates I would like to examine.  First, even small amounts of sea level change can drive changes to flood frequency, but it isn't immediately obvious that is what is going on here. The second option is a cool tidal phenomenon called the lunar nodal cycle, that I explore a bit here, or oceanographic anomalies drive by ENSO, may be partly responsible. I'm afraid it will take a bit of digging to really work that out. 

The transformed beach that supported the landing at Canoe Journey 2025

Yesterday I had the immense good fortune and honor to be on the beach as some of the first canoes came ashore for this year's Canoe Journey, hosted by the Lower Elwha Klallam Tribe. The theme of this year's journey, "A River Reborn" is a great one, but in my mind doesn't quite convey the full scope of the restoration associated with dam removal, only because when most of us think of a river, we think of the part of the river that flows downhill. But the Elwha River restoration is a great reminder about how connected those rivers are to the surrounding shoreline and coastal zone. To try to convey that point I want to focus on the place that canoes came ashore, here, at a point around 1000 m east of the river's outlet into the Strait of Juan de Fuca. This spot on the beach is very much part of the story of transformation associated with dam removal.  

August 2011 photo of the Canoe Journey landing area

Back in 2011, a month before the dam removal started, the lower part of this section of shoreline was very cobbly, with a narrow sand and gravel beach backed by a vertical eroding scarp. It would not have made for a good place to land canoes that can weigh, in some cases, up to 2000 pounds.  That same section of beach now has been utterly transformed by sediment released by the dam removal.

May 2025 photo of the beach at the Canoe Journey landing area

And while landing a huge canoe like the ones that came ashore yesterday cannot be easy, the transformed beach, I think it is fair to say, made it just a bit easier.

We've told the story of the response of the Elwha shoreline writ large to dam removal elsewhere, but here are some before/after beach profile and shoreline position data specifically from the Canoe Journey landing spot:

In the plot at top the black line is my most recent survey data from that location, collected in May. The orange line is a profile collected in August 2011, just a month before dam removal started.  The beach is notably higher, built up by additional sediment from dam removal, and also shifted seaward by something like 50-100 feet.  The shoreline position time-series at bottom is also interesting, as it shows that the beach continued to erode after dam removal started, until about mid-2014, after which there was a period of significant seaward beach growth.  Since that time this part of the beach has wiggled around a bit but largely held its position. 

August 2011 beach substrate photo collected at about +3 feet elevation on the beach

Beach profiles, though, only show us the shape and position of the beach, and don't tell the whole story of the transformation.  Back in 2011 this beach was also relatively coarse across much of the beach face.  The photo above, for example, is from August 2011 and was collected at just about the elevation of the beach that canoes were nosing ashore yesterday was I was watching around 10am.  This same elevation on the beach now looks like this:

May 2025 beach substrate photo at about +3 feet elevation on the beach

This is the transformed, and much more pleasant, surface that paddlers were able to leap on to as they brought these big canoes ashore:

A huge thank you to the Lower Elwha Klallam Tribe for the on-going support for the various monitoring and research project that I do on the shoreline, often with colleagues from USGS, Washington Department of Ecology, UW and others.  Also, thank you for the opportunity to volunteer at this year's canoe journey, and the opportunity to observe the landings yesterday.  

A Kalaloch state of mind

 

Manual panorama of the bluff edge at Kalaloch Lodge, 2 September 2016

I've been spending a lot of time of late thinking about Kalaloch. The main reason is that over the past few years I've developed and delivered a whole series of talks on Kalaloch, culminating most recently in a full one-hour presentation as part of this winter's Olympic National Park Perspective Series (that is available to watch here). That built on a few previous, more science-y sorts of talks presented at NPS Science Days in 2024 and 2025, the first of which is available here. That sort of thing always forces me to try to dig into a place a bit, and ring as much of a story as I can out of whatever insight I can find.

The issue at Kalaloch, or at least the issue that consumes my thinking, is erosion of the coastal bluffs there.  In the two panoramas above, taken from about here in 2016 and 2025, you can get some sense for this erosion. Back in 2016 the bluffs in this area were just at the tail end of a period of relative stability and were covered with vegetation.  Over the subsequent decade erosion in this zone just to the north of Kalaloch Creek appears to have accelerated, and the bluffs are now actively eroding, and nearly vertical.  

User-submitted photo from Chronolog station OLY-102 located here, taken 29 February 2024

That rapid erosion has had some significant consequences, including impacts to the trails and stairs used to provide access to the beach, as well as threats to bluff-top cabins rented through Kalaloch Lodge. Olympic National Park, in collaboration with Delaware North, the current concession-holder at Kalaloch Lodge, have instituted a cabin removal program to ensure that at-risk cabins are not occupied, and don't present a risk to human safety. The strategy is quite simple - when the bluff edge comes within 5 meters of a cabin, the cabin is closed and slated for removal. The cabin in the photo above, for example, was removed just a few weeks after that photo was taken:

User-submitted photo from Chronolog station OLY-102, taken 22 March 2024

I think there is more to come from the work that I and others have been doing to try to understand bluff erosion at Kalaloch. In addition to the monitoring that I've been doing at the site for more than 10 years, we also now have our five Chronolog stations at the site, that have been heavily used by visitors and are archiving a record of conditions at Kalaloch that we've only just started to fully utilize. The current erosion has also prompted some examination of the history of the site, which suggests that change is nothing at all new at Kalaloch. The map below, for example, was put together by Olympic National Park, and overlays a 1940s era site map on a more modern aerial photograph of Kalaloch:

There is a lot going on here and it takes some time to digest, but provides some  sense for how much the infrastructure on the site has changed through time (a story that is told admirably by David Emmick in a number of his books about the coast), and also how much the bluff has changed in the last 50 years (the 1974 bluff edge on the map comes from shoreline change assessment work done for Olympic National Park by the Washington Department of Ecology, published in 2002). But one thing that struck me about the 1940's map is what appears to be a sort of embayment where the Kalaloch Creek channel is now. Some of the early historical references to the site talk about "Kalaloch Cove", and this map suggests a waterbody that suits that name used to be there. In my mind this also hints at even more profound changes to the shoreline at the site than what we are seeing now. That is compelling to me, and I'm hoping to be able to continue thinking on the evolution of the shoreline at Kalaloch.

The non-tidal residuals of December

Flooding in Westport, Washington on December 14, 2024

This past December has been notably stormy, including a few events, especially on the coast (photo above) and along the Strait of Juan de Fuca (check out these great aerials from Three Crabs Rd near the mouth of the Dungeness River by John Gussman at DC Productions), that led to some flooding and other damage. The highest water levels of the month occurred on December 14th and were associated with a storm coinciding with one of this year's winter "king tides", But overall, the month has been characterized by a lot of positive "non-tidal residual" (referred to hereafter as NTRs), which is the part of coastal water level controlled by weather and not by the astronomical tides. NTRs are recorded by tide gauges, and for our purposes here are simply defined as the difference between the predictable astronomical tide, and the actual measured water level at a tide gauge.  

The non-tidal residuals recorded by the Friday Harbor tide gauge in December 2012 and December 2024

Back in 2012 I wrote here about a similar December, during which NTRs were generally high for a good bit of the month.  To illustrate this the plot above compares the NTRs from December 2012 and December 2024 as recorded at the Friday Harbor tide gauge. The things to note here are 1) that those NTRs were positive for long stretches of both months - in other words the measured water level at the tide gauge was higher than predicted most of the time and 2) that the maximum NTRs exceeded 0.50 m, or roughly 1.5 feet, on multiple occasions in each month.  For our region those are relatively large NTRs.

The manual staff gauge used by coast watchers on Puget Sound to measure water level

But what I want to focus on a bit here is a bit of a rabbit hole I went down because of an email discussion I had the good fortune to be part of with a group of property owners on Puget Sound. They reached out to discuss some of their water level forecasts and observations, and I absolutely love these kinds of notes, and the thought that there are people out there that really dig into this kind of stuff. This particular group has devised a really savvy system for both measuring (using a manual staff gauge, photo above) and predicting extreme water levels on the coastline so they can prepare for flooding along their waterfronts. Amongst other things they were looking to discuss a few days in November during which their observations and predictions didn't really line up because of the difficulty in predicting NTRs.

NTRs (top) and air pressure (bottom) measured at three tide gauges in Washington over December 2024.  The weird measure early in the month from Friday Harbor is a low-quality measurement that will eventually be removed from these preliminary data, downloaded from NOAA.

The thing about NTRs in Washington is that they generally are well correlated with atmospheric pressure - so high pressure generally leads to low, or negative NTRs (i.e. measured water level is near, or slightly below predicted), and low pressure generally leads to large NTRs, such that coastal water level is higher than predicted.  Above and below, for example, are the NTRs and pressure measurements from three tide gauges in Washington, that show this relationship, referred to as an "inverse barometer" effect. The group that I was emailing with used this relationship in their forecasts, to predict the water elevation for their area by combining the predicted tide with atmospheric pressure forecasts.

NTR vs. Pressure for data collected at three gauges in Washington during December of 2024

But what this group pointed out in our email exchange were time periods when this relationship sort of broke down.  In particular they pointed to a day in November...November 21st, 2024, when this general relationship between pressure and the NTR didn't hold, and diving into the data I think I'm starting to see what they are talking about.  Here, for example, are the pressure vs NTR data for three tide gauges in Washington for November 2024, with the measurements from November 21 highlighted in red:

What is interesting here is that on this day there really is no relationship between pressure and the NTR...the pressure more or less stays the same whereas the NTR at each of these tide gauges is varying by over 0.3 m, or about a foot, over the course of the day. I don't have a great, clear way to explain this, except that it does point to two important lessons.  First, the ocean is messy, and clean relationships almost never hold all the time. While there clearly is some relationship between atmospheric pressure and NTRs when looking at the data in aggregate, when we focus in on this particular day (and there are others like it), the relationship isn't so clear. Second, there are all sorts of other processes that can lead to variations in NTRs. My friend and colleague Eric Grossman at the USGS, for instance, has explored this problem in a paper, describing what they call remote sea level anomalies that are generated in the Pacific Ocean and can propagate into the Salish Sea. While it is not completely clear if the processes that are alluded to in this paper explain these observations, at least its a place to start.

Anatomy of a costal storm: January 9th 2024

 

As the winter storm season closes in on us on the coast of Washington, and on the eve of what is shaping up to be a pretty significant storm, I wanted to spend a few minutes looking back at the anatomy of a pretty interesting coastal storm that hit the coast and western and central Strait of Juan de Fuca last winter...on January 9th 2024. This one was notable to me because it was pretty damaging across the Port Angeles waterfront, and ended up shutting down Ediz Hook, a long, sand spit that forms Port Angeles Harbor that I spend a lot of time studying, surveying, and generally thinking about. 

A view along Ediz Hook NOT taken during a storm...April 2012

While Ediz Hook used to be a relatively low sandy spit one of its distinctive features now is a huge rip rap defensive breakwater on its seaward side, part of an Army Corps of Engineers Erosion Control project built in the 1970s. This rip rap raises the elevation of the seaward crest of Ediz Hook to elevations of around 9 to 11 feet above MHHW...pretty darn high...and up until this January 2024 storm I had never observed or heard of anything close to over-topping of this feature.  But this storm did it, sending water and debris over the breakwater, flooding the road that runs along Ediz Hook, and ultimately closing Ediz Hook until the storm subsided and crews could clear the road.  

Flooding of Ediz Hook on 9 January 2024

So what were the ingredients? Like most of these stories it starts with a winter high tide, predicted at about 8.5 feet MLLW, just a shade shy of highest astronomical tide at this site (coming in at 9.2 feet).  Also like most of these stories, that high tide was bolstered by a storm surge of about 1 foot, bringing the observed still water level up t 9.5 feet MLLW.  This is a high tide, but well clear of the record of 10.5 feet MLLW reached in 2003, and not even close to making the top 10 highest water levels recorded at this station since observations started in the 1970s.  

Driven by the high tide and storm surge alone, this storm was a nothing. But layered on top of this high water were waves...but interestingly the wave story has a similar narrative to that for tides and surge: The waves were big, but nowhere near record-breaking.  A wave buoy just to the west of Port Angeles, for example, recorded significant wave heights of just about 10 feet, whereas the New Dungeness buoy in the Strait of Juan de Fuca recorded significant wave heights of about 8.5 feet.  These are big waves, no doubt about it...but nowhere near maximums.  The New Dungeness buoy, for example recorded a maximum significant wave height of over 12 feet during a storm in December of 2006.  

So, like many of the impactful storms on the coast, this one was about a variety of factors coming together and co-occurring...in this case the tide, storm surge and waves.  None of those factors was necessarily extreme in isolation but taken together they created something powerful and noteworthy. This kind of interaction amongst various factors, none of which are extreme or present risks in isolation, poses real challenges in understanding these types of storms and their impacts, accurately forecasting them, and even building a sense for how likely they are to occur in the future. 

Buried silt on the Elwha sea floor

This week I was, once more, part of an annual subtidal survey of the marine community around the Elwha River delta during which we visited our site called, 4SP1, just to the east of the Elwha River mouth, roughly here. This is a place that was buried under 2-3 feet of sand in 2013, as the dam removal pumped out tons of sediment into the Strait of Juan de Fuca.  The site IS still very sandy, here is a bit of a view from this year:

and it does still look very different than it did before dam removal started (check it out in 2011 here).  But this year the site was also obviously different from last in that much of the deep mantle of sand deposited during dam removal had been stripped off.  So that is interesting in and of itself...we've been curious about how the Elwha River delta would evolve in the long-term now that dam removal is long over, and this gives us some indication.  But what really struck me was what was UNDER the remaining thin layer of sand on the seafloor...a few inch thick layer of silt and mud, which you can see in this photo below, with my slate for scale:

And get a sense for the composition in this video.  So what is the big deal?  This is a high current site, where mud and silt generally doesn't stick around.  I think this layer of mud and silt tells us something about how rapid and how massive the delivery of sediment was during the dam removals.  I envision a pulse of very fine material blanketing the sea floor, presumably followed almost immediately (perhaps within a tidal cycle?) by a pulse of sand, both pulses massive enough to blanket the seafloor with many inches of material...