Wednesday, November 13, 2019

Slow Start to Winter Snow Season

After the extraordinary early snow during the first half of October, many skiers and snowboarders were greasing their skies and prepping their gear for an early start of the snow recreation season.

But alas, things did not work out that way.  Warmer temperatures followed and then starting October 25th, we entered a 14-day dry streak.  Most of the snow melted, leaving us worse than a year ago.  To illustrate, compare the snow depth today and one year ago (see below).  The north Cascades and southern BC had much more snow at higher elevations in 2018.


So what about this year?  We have had a "problem"--persistent high pressure over the northeast Pacific.   To illustrate here is the height anomaly (difference from normal) at 500 hPa (about 18,000 ft) for the past 30 days.  Red indicates higher heights (higher pressure/ridging), blue indicates troughing (lower than normal heights).  A strong persistent pattern that explains the dryness and warmth of the West Coast and the cool/wet weather of the eastern US, since ridging brings drying and troughing, the opposite.

The forecast for the next five days for the same level?  A big ridge over the west coast (the shading is the anomaly, with orange being higher than normal).  This is a wetter pattern, but relatively warm.  Not good for snow.

The latest 7-day accumulated snow forecast from the UW system provides some snow at the highest elevations over the north Cascades, but even there only 6-12 inches.

Strangely enough the persistence of this ridge could be a good thing for snow later in the season--if the ridge moves back towards the west, we might be open to a trough coming from the north, like last February.  We will see. 

The media has been full of winter weather forecasts, the skill of which is marginal at best.

Sunday, November 10, 2019

Can Autonomous Weather-Observation Sailboats Improve Forecasts over the U.S.?

Imagine an unmanned sailboat that can collect highly accurate weather and upper-ocean observations.

A sailboat that could stay offshore for months or more and can handle the most severe weather.

A sailboat with real-time satellite communication and capable of either staying at one location or following a pre-programmed path.

A sailboat observation platform that is not only more capable than current weather buoys far less expensive to deploy and maintain.

Now imagine a line of such sailboats stationed off the entire U.S. West Coast, providing a "picket fence" of weather/ocean observations that would help warn the U.S. West Coast of approaching weather systems.  Observations that might improve weather forecasting over the entire nation.

Too good to be true?  Science fiction?   The product of too many meteorological happy hours?

No, this is now reality.    The observing sailboat is called a Saildrone, produced by the eponymously named company and the brain child of its visionary founder and CEO Ricard Jenkins. 

A line of Saildrones is now off the West Coast as part of a joint Saildrone/ University of Washington project:  The Saildrone Pacific Sentinel Experiment.   And the implications of this new observing platform are potentially revolutionary.

The Saildrone

   The figure below illustrates the standard observing systems of a Saildrone.  For the atmosphere, all the old favorites:  wind speed and direction, temperature, humidity, barometric pressure, solar radiation. For the ocean, everything from ocean temperature below the surface, the temperature of the sea surface, wave heights, and near-surface ocean currents--to name only a few.    Plus, the Saildrones have multiple cams, so one can view the sky and sea surface. Amazing.

The "Competition"

            NOAA/NWS (National Weather Service) runs a major buoy program, with the weather/ocean buoys off the U.S. West Coast shown below. The blue squares are the fixed weather/ocean buoys.  There are five of them offshore.  Three of them have either failed or have 3 or more critical sensors broken.   The other two, not so well placed, have one important sensor gone.   Not good.  The red stars are tsunami buoys. The density of the fixed buoy network is relatively sparse, and the buoys tend to break or get loose from their moorings after major storms. 
            The other problem with the NOAA fixed buoy program is its costs: hundreds of millions of dollars a year, including the need for very expensive ship time to service, retrieve, and deploy buoys.

The offshore buoys are important for West Coast forecasting, particularly the buoys at 130W.  They tell us how well our forecast models are doing as storms approach the coast and whether the forecasts need to be updated or amended.  They supply data that can be assimilated into models to improve downstream forecasts.  Unfortunately, NOAA does not have a functional line of buoys offshore.

The Saildrone Sentinel "Last Line of Weather Defense" for the US West Coast

As someone who has spent his life working to improve weather forecasts along the U.S. West Coast, I have often been frustrated by the unfortunate state of the offshore NOAA buoy network.  My colleagues in the National Weather Service feel the same way. Although West Coast weather forecasts have improved, with better models and more satellite observations, some major forecast busts have occurred, such as the (in)famous Ides of October storm of 2016;  we have lacked sufficient surface observations offshore to view the failure mode of this storm and other in a timely way.

I had heard about Saildrones and was intrigued by their potential to improve weather forecasting.  So intrigued I contacted, Richard Jenkins, CEO of Saildrone.   Graciously, he invited me down to his impressive facility in Oakland, where I saw a dozen or so boats being constructed and talked about their use as a data source for numerical weather prediction.

I was excited enough about Saildrones that I wrote a white paper describing a project in which a line of Saildrones was positioned off the West Coast, about 300-400 miles offshore.  A "picket fence" dense enough to get a piece of any major system approaching from the west.   The graphic below shows what I proposed to Mr. Jenkins:  six Saildrones (for a start), indicated by the yellow pins, with the red ones indicating currently operational NOAA buoys (but with some sensors broken).

Richard Jenkins was enthusiastic about the project and during the summer his company prepared the boats and some moved into position during September and early October.  Five of the boats are now in place as I write this and providing us with real-time atmospheric and ocean information.  The last boat will head out to sea this week.  A UW team is now in place for evaluating the potential for this new observing platform (including my colleague Greg Hakim, who specializes in data assimilation, research scientists Jeff Baars and Rick Steed, who have built the data acquisition and display infrastructure and will work on modeling experiments, and  UW scientists David Ovens and Robert Tardif who will help with modeling.

To allow you to view the Saildrone data in real time, members of the team have created a project web site (  You can view the new observations, how the observations differ from what NOAA/NWS thought was out there, and much more.

A sample of the cams on one Saildrone

An Adaptive Observing System?

Saildrones obviously can move (as much as 50 miles per day, depending on the wind).  Could we take advantage of this capability to improve observations of important storms?    Weather forecasts are now quite good 4-5 days out.  Let's say a storm is going to approach the central California coast, but there is some uncertainly of exactly where.  With that much time we could move "picket fence" of Saildrones to greatly increase their density in the area of uncertainty, and thus hopefully, lock down the storm characteristics, improving downstream forecasts.  Or perhaps a hurricane is forecast to threaten the East Coast, but with great uncertainty in location (such as with Dorian).  In such case, the source of the uncertainty is often off the West Coast of the U.S.  We have techniques for determining the key areas that observations are need and could move the Saildrones into position.

A Replacement for the NOAA Buoys System?

Currently, NOAA fields a little over 100 fixed buoys.   Many have stopped working, are adrift, or have major sensor failures (current status here).  Winter storms cause major damage to these buoys.  The only way to fix or retrieve them is through very expensive ship operations.   But now imagine replacing them with Saildrones.   Problems with a Saildrone?  Just sail a replacement out there!  And Saildrones could be adaptive, moving around for optimal positioning.   I suspect that this approach would save millions or tens of millions of dollars.

Sunday afternoon Saildrone Observations

The Project

Here are some of the big questions we would like to answer this winter:

1.   Can Saildrone observations offshore delineate deficiencies in the NWS atmospheric analysis over the eastern Pacific?
2.   Can we improve numerical weather prediction by using the Saildrone observations?
3.   Can we secure a better picture of approaching major storms (e.g., Pacific cyclones, atmospheric rivers, etc.) using this new platform?
4.  Can Saildrones be shown to be a robust observing system for the very stormy northeast Pacific?
5.  Can we learn about offshore weather and ocean features using this new platform?
6.  Might we might be able to adjust the position of Saildrones based on forecasts to better observe important storms?

The first stage is now done:  building the software infrastructure to access and display the Saildrone observations and to compare them to NWS analyses.   The next step will be to complete numerical weather prediction experiments for a limited number or cases (particularly ones in which the Saildrones were suggesting that the NWS model analysis were problematic.

This is going be very exciting.   Not only the science, but the potential to improve U.S. weather prediction for major weather events.  Much of the U.S. is downwind of the Saildrones and potentially we could improve forecasts for the lower 48 states.  And the use of Saildrones might be able to save NOAA millions of dollars per year by replacing the pre-existing buoy network.

Friday, November 8, 2019

Will You See the Transit of Mercury on Monday Morning?

A relatively unusual astronomical event will take place at sunrise (here in the Northwest) on Monday morning:  a transit of Mercury across the Sun.

The last time Mercury cross the sun was in 2016, and the next time will be 2032.  Here in Seattle, the transit will begin at 7:08 AM (sunrise), be at the midpoint at 7:20 AM, and end at 10:04 AM.  Total duration will be 2 hour, 56 minutes and the transit will be already started when the sun rises.

The big question is clouds.   Let's examine this issue by looking at the UW WRF cloud forecast for 7 AM Monday morning.  The model is predicting a middle/higher level cloud deck, albeit thin over our region (the gray color).

A simulated infrared satellite image, based on the model output, is shown below. Thin high clouds.

The bottom line is that viewing of the transit is uncertain.  If the cloud veil is thin enough, perhaps.   But the transit is a subtle feature--a small dot moving across sun's disk.

And we have another nail biter tonight.   If it doesn't rain before midnight, Seattle will tie the record for dry days in November (14th).  But the radar shows the frontal precipitation approaching offshore at 9 PM (see below).  Fingers crossed.

Wednesday, November 6, 2019

One of the longest November dry periods in Northwest history

I am sure it has not escaped your attention that it has been generally dry and sunny the past few weeks over the region, a situation that is becoming increasingly unusual as it extends into our normally wet/stormy November.

By the end of today (Wednesday), we have gone through 12 days without measurable precipitation in Seattle (the last day of precipitation at Sea-Tac was October 25th).

To get an idea of how unusual such a dry streak is, consider the following graphic produced by Dr. Joe Zagrodnik, which shows that the longest dry streaks since 1950.  For the November period, the longest dry streak on record at SeaTac is 14 days.  12-days is tied for fourth place.
But our dry spell is not over yet.

Let's take a look at the ensemble forecast (models run many times, each a bit different) of accumulated precipitation at Sea-Tac Airport from the NOAA/NWS GEFS modeling system (see below).  None of the forecasts have any precipitation through 4 AM on November 9th (Saturday).  So two more days are in the bag.  We will tie the record dry streak of November (14)!  Enough to get weather enthusiasts excited.

But we have a real chance, but not a certainty, of completely busting the record.  The high resolution member of the ensemble (blue line) has no precipitation through November 14th as does several of the ensemble members (gray lines).   More members show some rain starting around November 13th.

What about the world-leading European Center ensemble (see below), which includes 51 members?  Dry through Friday.  We will tie the record.  Saturday is on the edge, with a number of runs with very light precipitation.  There really is not much precipitation through 2/3rds through the month.

So enjoy the next few days.  And perhaps we might take on another record--the lowest November precipitation---but that is one I would not bet on at this point in the month.

Monday, November 4, 2019

New Generation Weather Satellite Provides a Clear View of Nighttime Fog

We are now in the middle of Northwest fog season and the conditions are nearly ideal for low-level fog and low clouds.    The nights are getting much longer, allowing good radiational cooling from the surface.   We had rain, so the surface is relatively moist.  A high pressure ridge is just offshore producing sinking air over the region (see figure).  High pressure is good because it suppresses upper level clouds and produces warming aloft (sinking air is compressed and warms).  High pressure also produces weak winds--which is helpful for fog, since strong winds mix the cold air away from the surface, lessening the potential for cooling to saturation (and thus produce fog).

Meteorologists would like to see the extent of the fog at night---but there is a problem.   Visible satellite imagery doesn't work at night, because, well, it is night.  Infrared satellite imagery works all the time by sensing the temperature of the clouds--but there is a problem regarding fog.  Fog is very shallow (typically from meters to a few hundred meters deep) and thus the temperature of the top of the fog is very similar to that of the ground.  This is particularly true of fog-prone nights when there is frequently an inversion (temperature warming with height near the ground) or little change in the vertical near the surface.

So how do we track fog at night from space?  It turns out that weather satellites look at the earth in many wavelengths (or channels), and the difference between one of them (3.9 microns, a micron being a millionth of a meter) and one of the infrared wavelengths (10.3 microns), can indicate where low clouds are at night.  This difference is called the Fog Product.   

We had fog products with the old NOAA/NWS geostationary satellite (GOES 15), but a year ago,  the new National Weather Service GOES satellite was launched (GOES 17) and its higher resolution is just a home-run regarding the clarity of the fog images.

Let me show you.  Late Sunday night, there was little fog/low clouds over the interior, but plenty offshore, with some of the densest fog along the coast.

By 1 AM, fog was forming along the I-5 corridor south of Olympia (an infamous fog hole) and along the Columbia River.  Some of NW Washington.

But around 6 AM today, fog had really expanded over SW Washington and the northern Willamette Valley,  Portland had lots of fog.  There was some high clouds over the NE part of the domain that was interfering with the fog product....hard to get around that.

Knowing the extent of fog is important for transportation (airports, highways), among other reasons, so knowing the distribution of fog is extremely useful. 

One of the first visible images this morning, showed the fog was a reality!

Finally, fog is generally not the best forecast field by our numerical models, but high-resolution models are getting better at it.   To illustrate, here is a 15-hour forecast from the UW, uber-high resolution WRF system for low clouds and fog valid at 7 AM.  Not too shabby.

Finally, remember that on cold nights be careful driving when there is fog on or near the road.  Fog can quickly lay down ice on any road that falls below freezing.

Saturday, November 2, 2019

Rain Returns to the Northwest

It has been extremely dry during the past week over the Pacific Northwest, and dry conditions will continue into Thursday.

But by the weekend, the rain will return. 

Dry periods extended into the first week of November is not that unusual in western Washington and Oregon, but for it to continue into the second and third week would be  extraordinarily unusual.   And this won't happen this year.

Let's forecast the right way, looking at ensembles of many forecasts.  Below are the 21 members of the  U.S. GEFS ensemble.   The accumulated precipitation of each member at Sea Tac is shown. Precipitation is flatlined for all members until November 7th, after which there is a steady rise.  This is called a plume diagram by the way, with each line being one member.

 What about the more skillful and larger (51 members!)  European Center ensemble?  A similar story, but presented a different way--a stripe for each forecast, with the colors indicating the 24-h precipitation amounts.  Neither system is indicating a  large amount of precipitation at Sea Tac.

The precipitation total through 4 PM Saturday shows the heaviest precipitation is going into southern British Columbia, but with several inches extending into northern Washington State.  Unfortunately, temperatures will be too warm to result in much snowfall in Washington.

Thursday, October 31, 2019

Extended Forecasts are Not Reliable

We are constantly exposed to extended forecasts in the media and online, with predictions extending through the next month and more.

Can you rely on such predictions?    Are they really worth paying attention to?

Quite honestly, probably not--and if you do consider them, do so with the knowledge that their skill is marginal at best.

Take this month (October) for example.  The official NOAA Climate Prediction Forecast for October temperatures, made on Sept. 19th, was for warmer than normal conditions over the west and MUCH above normal over the southwest U.S.

What actually happened?  Nearly the entire west was much colder than normal, with the northern parts MUCH, MUCH colder than normal.  A miss.  In fact, a big miss.

Or the official 3-4 week forecast, made on October 4th?   Warmer than normal over the west.

Such poor forecasts even a month out are not unusual.   UW graduate student Nick Weber and I evaluated the skill of the main U.S. long-term forecasting model (the CFSv2) and found that skill is typically lost after roughly 2 weeks (see below and published in the peer-reviewed literature).  This figure shows the forecast error (root mean square error) at 500 hPa---about 18,000 ft, a good level to view atmospheric predictability.  The situation is the same over Washington, the western U.S., the continental U.S. or global.  Skill is rapidly lost the second week out.

While meteorologists struggle to produce improved forecast skill past two weeks, we have gained a great deal of skill at the shorter time ranges, particularly for days 3-8.  

So why is our skill improving rapidly for the shorter periods, but not the longer ones?

Because the forecasting problem is very different at the different temporal scales.

For the short periods, forecasting  is an initial value problem.  We start with a good description of the 3D atmosphere and our models simulate how things evolve.   Because of weather satellites and other new data sources, our initial description of the atmospheric has gotten MUCH better.  And our models are much better:  higher resolution, much better description of key physical processes, and more.  That is why a plot of the skill of  skill of the 1-10 day forecasts of the European Center has improved greatly over the past decades (see below)

But small errors in the initial description of the atmosphere and deficiencies in our models inevitably lead to growing errors, and by 2 weeks such errors swamp the forecast.  The forecasts are not much better than simply using the average conditions (or climatology).

There is hope for some skill beyond two weeks, by taking advantage of the forecast skill available from aspects of the environment that are changing slowly (such as sea surface temperatures, sea ice extent, snow extend, soil moisture).    These aspects influence the atmosphere and potentially can torque the atmosphere one way or the other.  Essentially, the forecast problem has changed from an initial values problem to a boundary-forced problem (the boundary being the surface characteristics that can influence weather).

But the skill that might be available from the boundary conditions is different---not about the conditions at a specific time, but for the average conditions over a month or season.   A good example of such skill is the relationship of the warmer (El Nino) or colder (La Nina) temperatures of the tropic Pacific sea surface and weather around the world.    There is some skill there, but it is relatively modest.  

Unfortunately, our models still have key deficiencies (such as poor description of thunderstorms) that make it difficult for us to derive all the potential skill that should be available from the slowly changing boundary conditions.   A lot of work is needed, but I am hopeful that eventually forecast skill beyond two weeks will improve.  

Tuesday, October 29, 2019

A Surge of Cold, Dry Air Envelopes the Northwest and Now Threatens California

Last night, very cold, dry air surged southward out of Canada, pushing into Washington and Oregon.

The pressures, temperature, and wind map for 8 AM this morning tells the story (below). The blue is cold air, with the primo stuff east of Cascades.   Note the large pressure difference across the Cascades, which resulted in strong easterly winds in the Columbia Gorge and in Cascade mountain passes.

If you are in Washington State, do you feel under pressure?  You should, since we have usually high pressure at the surface right now.    For example, at Ellensburg, in the central portion of the State, the sea level pressure is higher than any other time in the last six months, reaching nearly 1040 hPa (mb).

And how about your skin?   Does it feel kind of dry?  It should, sicne e have some of the driest air in a very long time over us--yes the driest since the start of the year--as indicated by the very low dew point temperature at Sea Tac Airport.    Dew point is a good measure of the amount of water vapor in the air (and it is the temperature at which air becomes saturated when cooled).  This morning it got down  below zero (in fact, -4F).   Mama Mia...that is super dry.  Even drier than the arid period last March when downslope winds started fires around western Washington. 

The large pressure difference across the Cascade resulted in easterly winds accelerating to 50-60 mph around Enumclaw and Tacoma today and similar winds around the Columbia Gorge and some of the highest hills in eastern WA (see below, click to expand).

As a result, there were some power outages in the south Sound today (some still going on)

With very dry air above, the greenhouse warming effect of water vapor in the atmosphere will be lessened, and our temperatures should plummet tonight.  Be prepared for lots of frost when you go out tomorrow morning--and some cooler areas away from water will drop into the teens.

This cool dry air will slowly warm, but expect sunny skies and cool nights for the rest of the week.

But the real worry is for the folks in California.  The cold air moving south has an Amazon Prime account and the delivery will be overnight.   By tomorrow morning, the cold, dense air and associated high pressure will have moved southward through Nevada, setting up a large pressure difference across the Sierra Nevada (see map below for 5 AM Wednesday).  This pattern will set up Diablo winds for northern CA and Santa Ana winds for southern CA, threatening both areas.

Southern California will face a real challenge with a powerful Santa Ana wind descending to the ocean by 5 AM Wednesday (see below).  The most severe winds (sustained 50 mph with gusts to 80 mph) will be east and north of downtown LA, but substantial winds will be felt southward down to San Diego. 
 The winds will weaken a bit by 5 PM Wednesday (see below), but will be more than strong enough to sustain any fires that started earlier.  It is crucial to turn off the power to a lot of folks in southern CA tomorrow. Turns out the Getty Fire was the result of trees on powerlines--just a taste of what is possible with the far more powerful winds overnight.

Sunday, October 27, 2019

Extreme Winds in California--and Soon in the Columbia Gorge-- Are Being Driven by Unusually Cold Air

It is perhaps a great irony.   

Winds gusting to 70-100 mph over northern California are contributing to a major wildfire and threatening more.  Nearly two million people have lost power.  The Columbia Gorge will soon experience hurricane-force gusts.

And it is all because of much colder than normal air moving southward east of the Cascades and the Sierra Nevada.

A Great Irony.  Colder than Normal Temperatures Produce This

To show you the cold air,  here is map presenting the differences from normal (the temperature anomaly) at 11 AM this morning.  Some locations in Nevada are16 to 24F BELOW NORMAL.

Another pulse of cold air comes in Tuesday and Wednesday.  Here is the forecast for Wednesday at 5 AM.   Makes me shiver to see it. Eastern Oregon and Nevada are 15-20 F colder than normal.  Idaho and Montana are 30-40F below normal. 

The colder than normal air east of the Cascades and Sierra Nevada results in higher surface  pressure there, since cold air is denser and heavier (surface pressure is a measure of the weight of the air above).  You can see the effect is morning (8AM), with the figure showing pressure (solid lines) and temperature (shading, blue and white are coldest).   Big pressure change over the Sierra and northern CA, producing strong winds.

You want to see something really impressive? is the pressure/temperature forecast for 8AM Tuesday.  Wow.  Mucho cold air moving into eastern WA, eastern Oregon, and northern Nevada. crazy cold over Montana.  Notice the large pressure gradient across the Cascades? That will produce very strong eastern winds in the Columbia Gorge.

 To illustrate, here is the gust forecast for 11 AM Tuesday around the Gorge.  Winds above 50 knots are predicted (dark blue color).

On Wednesday morning (shown below), the cold air will push southward, setting up ANOTHER major Diablo wind event over northern CA and a MAJOR Santa Ana over LA.

Now several folks are talking about this event being associated with global warming.  CA Governor Newson said this in a press conference.    This is simply not true.  The vegetation is dry, like it always is, this time of the year.   And even if the grasses and bushes were wet, the air is so dry that the grass would be ready to burn in hours.  

How dry?  In several locations this morning, the relative humdities got down UNDER 5%.  Lots of places under 10%. Unbelievable.  This is due to the descending easterly flow associated with the current Diablo wind event.

During the past decades, cold air outbreaks have weakened, partially due to global warming.  And long term climate models show that global warming will increasingly weaken the cold air outbreaks that drives the strong easterly winds that drive CA wildfires in the autumn.  Global warming has all kinds of negative effects, but one of them is NOT the strong easterly flow that drives the big wildfires we have seen in recent years.