Wednesday, May 20, 2015

Is it still looking wet for this Memorial Day Weekend?

So a couple days have passed since our last blog post about the weather outlook for the Inland Northwest for the upcoming Memorial Day weekend. Consequently, our confidence in what sort of weather to expect should be increasing. Recall that in the last post, there were suggestions that this could be a very wet weekend. In fact, perhaps the wettest Memorial Day weekend since 1997 and the second wettest on record (since 1970). So is that still the case?

Saturday 500 mb height
Above is the 500 mb map for Saturday from the GFS. This is a similar setup to the maps we explored earlier this week with a deep low centered over the Pacific Northwest. So is this GFS run an outlier or do other model agree with the solution?

4 different models showing 500 mb heights for Saturday (upper left model did not come in fully)
Here's is a look at 4 models compared to each other and they all show a low focused somewhere over the Pacific Northwest on Saturday. So where is this low on Sunday?

500 mb maps for Sunday
Again there is great agreement that there will be a low focused somewhere over the Pacific Northwest. The top 2 models take the low into southern Oregon, whereas the bottom two models drop the low into southern Oregon. Now, what the positioning of the low on Memorial Day?

500 mb maps for Monday
Again all the models keep the low over the region, however, the consensus is to drop the low south of the Washington/Oregon border.

So our confidence in this event is quite high. We are certain this low will form and move over portions of the Inland Northwest however how long it will remain over the area and where its precipitation  band will set up is the big question. Looking at the GFS for the Saturday and Sunday yields a significant northwest to southeast orientated band of precipitation.

48 hr precipitation amounts for Saturday-Sunday (latest model run )
This is quite similar to the previous model run we looked at a couple days ago(see below). Both model runs show a similarly oriented band of precipitation, however,  the latest version has the band a little farther to the west leaving places such as Sandpoint and Colville much drier than the previous forecast.

Rainfall forecast for Saturday and Sunday (Monday's model run)
Obviously wherever this band sets up will determine who gets a lot of rain and who doesn't. If we decide to consult another model solution it also shows a similar NW-SE band setting up, but this time it's displaced even further west. If this solution were to pan out even Spokane would miss out on the bulk of the rain as would the Silver Valley. 

Another Saturday-Sunday 48 hr precipitation forecast

So given these uncertainties it will be hard to pinpoint exactly where the band will take up residence. This is where ensemble model forecast can sometimes help to pinpoint the band (s) of heavy precipitation. So once again referring to the GEFS/NAEFS ensemble forecasts we see that it maintains a NW-SE oriented band of heavy precipitation as well. However the model has backed off from previous runs which suggested this would be a 99-99.9 percentile event for this time of year. Nonetheless the model still suggests this is potentially a 90-97 percentile event which is still significant. 

NAEFS/GEFS Ensemble 48 hr rainfall Sat-Sun

So what about the temperature forecast? We are holding onto our thoughts of hitting near 70° each day in Spokane this weekend, however if the band hits our area what will our temperature be? To say there is some uncertainty in this forecast would be an understatement. The 12z run of the GFS (dark blue line) has our high hitting the mid-70s that day. How about the 18z run of the GFS? A chilly mid-50s! That is about as uncertain as things can get. Its pretty much the same story for other parts of the Inland Northwest including Moses Lake (bottom meteogram)

Meteogram for Spokane
Meteogram for Moses Lake

This forecast remains a problematic one. Suffice it to say a good portion of the Inland Northwest is going to see a soggy beginning to the holiday weekend. Specifically where that will be, we still can't say with good confidence.  We expect the wettest day of the three-day weekend will be Saturday with a gradual improvement expected after that.

Monday, May 18, 2015

How is this Memorial Day Weekend looking weather-wise?

Last year at this time, we composed an extensive blog discussing the connotations associated with Memorial Day weekends in the Inland Northwest.  In many parts of the country, this weekend typically kicks off the summer season. However here in the Inland Northwest that really isn't the case, due to continued cool and unsettled weather. So is this year going to fit the weather mold, or will we see warm and dry weather?

Before we delve into this year's weather scenario, lets look at what the typical weather pattern is for this time of year. Looking at the mean 500 mb (18k ft MSL) maps below we see that the typical weather pattern is to see moist southwest flow pointed into the Pacific Northwest with a ridge setting up just downstream of the Continental Divide in western and central Montana. 

500 mb mean map 5/23-5/31
So what about the temperatures associated with this average pattern? Here's  the average temperatures for a couple locations as well as the average chance of precipitation:

So not bad. An average high around 70°F with rain expected on roughly 1 out of 4 days doesn't sound too terrible. But that's climatology and this weekend's weather could vary significantly from the norm. Why you ask? Well, let us take a look at the 500 mb pattern for this weekend to get an idea what sort of weather we might be dealing with. First off, here's a look at the weather for Saturday. 

GFS 500 mb heights for Saturday
This does not look promising if you enjoy plenty of sunshine and warmth. Note the closed low-pressure system (yellow L) centered over northwest Washington. Low-pressure systems like this are typically associated with cool and unsettled weather, especially if your are under or near the low. But this is only one model run. Surely there are other ideas out there in weather land. Well, here's a look at 4 distinct weather models for the same time period.
500 mb pattern from 4 different weather models for Saturday
Notice all these model runs place a low somewhere over the Pacific Northwest. That raises our confidence that the weather will be cool and unsettled over portions of the region, but maybe this unsettled weather will only last for a day. So let us take a look at the Sunday weather pattern.

500 mb pattern from 4 different weather models for Sunday
If the upper left model solution verifies (the GFS), the Inland Northwest could really be stricken with another cool and wet day. The other model solutions would hint at some modest improvements, but far from what we would deem summer like weather. How about Monday then?

500 mb pattern for Monday
The upper left solution (the GFS) still keeps the low parked over Washington while the others either drop it south of the region or weaken it. We shall see. But what if the GFS solution were to verify? What would our weather be like? If you love rain, well then you will love this forecast. Here's a look at the total precipitation forecast for Saturday and Sunday.
Rainfall forecast for Saturday and Sunday
The swath of red shading indicates rainfall amounts in excess of an inch. This includes Spokane, Kellogg, Coeur d'Alene, Pullman, Lewiston, Banks Lake, Lake Roosevelt, and Colville. If you look even closer there are a few locations which are expected to see in excess of 2 inches (shaded in tan)!
Now that is a lot of rain, regardless of the time of year. If the precipitation were to materialize, it would equate to our wettest Memorial Day weekend since 1997 and the second wettest on record (since 1970). So what is our confidence level in this precipitation panning out at this point?

Another tool we can look at in addition to just a few deterministic model runs is from an ensemble forecast approach. This is where you look at multiple model runs (initialized from the same model) and try to gather where the forecast confidence in the precipitation swath is greatest. Using this approach, the news isn't great either.

NAEFS/GEFS Ensemble precipitation amounts for the 48 hrs ending Sunday Afternoon
Without delving too deep into the specifics of the image above, the black lines depict the average precipitation amounts from the ensemble mean precipitation forecast while the green and blue shading (well any shading really) is bad if you like dry weather.  Notice the black lines hint at a swath of precipitation extending from southern British Columbia to the central Idaho Panhandle. The green and blue shading suggests the 48-hour precipitation amounts are in the 99-99.9% percentile for an early April-early July event. In other words, it is a 1 out of 100 or 1 out of a 1000 event for this particular time of year! There is even a small area of pale blue which hints that rain of this magnitude for this time of year has not been experienced since between 1985-2011!  This map suggests the GFS isn't entirely alone in forecasting a wet weekend, but we would like to see much better model agreement before buying off on this solution.

Now, if the rain does materialize, what impact will  this have on our temperatures?
7-day Spokane Forecast
Our current forecast is showing highs above average through the workweek with a cooldown by the weekend. Notice by the weekend we are forecasting highs right around 70 each day. That's not bad. Right around average actually. But we discounted the wet GFS solution. What if the GFS were to verify? How warm would it get then?

7-day temperature meteogram for Spokane
See the green line on the chart above? That is our official NWS forecast. However the blue lines (from the GFS) show the highs on Saturday could struggle to escape the mid-50s on Saturday and then slowly rebound into the lower to middle 60s for Sunday and near 70 on Monday.

Stay tuned to later forecasts this week to see how wet this weekend might really get. In the meantime, you might want to procure yourself one of these if you have camping plans this weekend.

Thursday, April 16, 2015

Warmer weather ahead

After one of the warmest March's on record (see previous blog entry here), April has been off to a cool start, with snow falling at times in the mountains, and even in some lower valleys at times.  Here is a look at the first half of April temperature anomaly map.

As the map shows temperatures have been running about 1 to 3 degrees below normal over the northwest.  Has the snow helped out our mountain snow situation?  Here is the latest map showing where we are at

The mountain snow pack has improved very little.  While the percent of normal values have increased slightly, they are still below 25% (dark red) or 25-50% of normal (red) for most locations across the Inland NW.  Several sites are at or below record low values dating back to the late 1970s or early 1980s.

The weather pattern is currently changing to a warmer one!  A strong ridge of high pressure is about to move in and will last awhile.  Temperatures Friday through Monday will be above normal, with the warmth possibly peaking on Tuesday.  Here is a look at a forecast model showing the general weather pattern for Tuesday.

500mb forecast map valid 5 am TUE from the 06z GFS Model run
The strong ridge will likely result in temperatures that are well above normal.  Here is our current forecast for high temperatures on Tuesday.

What about after that?  The ridge may weaken and allow for some cooling although there is quite a bit of uncertainty for the middle to end of next week.

What about May?  CPC just released their new monthly and seasonal outlooks this morning (typically issued the third Thursday of every month).

The May outlook is calling for a good chance for warmer than normal temperatures.

What about June, July, and August?

Looks pretty similar, with elevated odds for a warm summer.

Friday, April 3, 2015

Will the Inland Northwest see snow on Sunday night and Monday?

The Spring thus far has been fairly mild, and as a result we haven't had to worry about the threat of snow. The temperatures for March were some of the warmest if not the warmest on record for many locations.

March temperatures
But all weather patterns come to an end at some point and this will be no exception. So here is what the upper-level weather pattern for most of the month of March looked like. It showed a rather persistent ridge focused over the West Coast with the storm track nudged well north of our area
Mean 500 mb pattern for March 1-March 29th
However since the end of the March and continuing into April, the upper-level ridge has shifted to our east, allowing the upper-level trough to nudge into the eastern Pacific and the storm track to push through the Pacific Northwest.
Mean 500 mb pattern March 30th-April 1st
This has resulted in more typical springtime weather for the region including significantly cooler temperatures, and numerous afternoon/evening showers and thunderstorms accompanied by an Inland Northwest springtime staple: graupel. But despite the cooler weather, most of the valley locations, have yet to deal with snow. That might change though with the arrival of the next significant weather system. The model solutions are rather convincing that this system will arrive sometime between late Sunday night and Monday as the deep offshore trough heads inland.

500 mb pattern for Monday morning
Referring to  the map above, it suggests most of the upper-level energy will remain poised off the southern Oregon/Northern California coast, however look closely over extreme northeast Washington/north Idaho and notice the small yellow circles. These indicate the larger trough will also contain several small shortwave troughs.  If these troughs are accompanied by enough moisture and instability they can trigger small bands of moderate to intense precipitation, or what we
meteorologists term meso-banded precipitation or mesobands. Even more intriguing is these mesobands can often allow snow to fall (and possibly accumulate) at much lower snow levels than expected.

The odds are quite good that the region will see mesobands of precipitation form, the big question is where and what will the impacts be?

To answer the first question of where will these bands form, we first need to look at the positioning and track of any surface low tracks. Here is a look at four weather models we typically utilize and where they place the low (and resultant precipitation) as of 5am Monday morning.

4 model solutions with surface low position and 6-hour precipitation valid 5am Monday
As you can see above, all four models have a surface low located over the eastern third of Washington or north Idaho, however, the exact positioning and strength of the low is quite variable. The solution in the upper left corner is by far the most impressive solution. It has the strongest surface low (1000 mbs) and the heaviest amount of precipitation. The purple colors in this image represent precipitation amounts between 0.50-0.75" in a six-hour period. The other model solutions show a weaker low (1005-1006 mbs) and much lighter precipitation amounts.

When there are wide disparities between our core models we like to defer to ensemble modeling. This is where we take an initial model run and add small perturbations to the mix. The perturbations begin small but with time tend to grow. When the models cluster the positioning of a low, this boosts our confidence in the forecast, whereas if lows are strewn haphazardly across the region, our confidence is quite low. So here's a look at the SREF positioning of the surface lows for Monday morning.

SREF Surface low positions for 5am Monday

Although there are plentiful lows (L's) found across the Inland Northwest there is a wide scattering of their positions. The fact that there are plentiful lows is good from a standpoint of there is fair confidence of an event occurring. However, the wide scattering of the L's lowers our locational confidence significantly.  We can also look at the mean of all the low positions and the mean of all the precipitation data to come up with a preliminary snow forecast. In this case, the SREF is showing this as the mean snowfall for the 12hrs ending at 11 am Monday.

Mean 12hr snowfall ending 11am Monday
While these values themselves are not impressive (1-2" for the areas shaded in green) the fact that this model is showing such widespread snow is noteworthy. Another interesting thing we can look at from the SREF model is a plume diagram which shows all the suite of model solutions on one chart. So given the map above, it would seem locations north and east of Spokane as well as near the Cascade Crest would see the best chances of snow. So to hone our forecast a little more we will look at some plume diagrams to determine potential snow amounts. The first we will refer to is for Sandpoint.

Sandpoint snow plume diagram
Each line on this plot refers to one particular member of the SREF ensemble. In this case, most of the solutions are showing some snow on the right side of the diagram. The mean of the runs is denoted by the black line which is indicating a mean snowfall of 2". More impressive (but not probable) is the pink line which shows snowfall totals exceeding 8".  How about the plume diagram for Spokane?

Spokane snow plume  diagram
Even the Spokane plume diagram is showing some snow. Nothing compared to Sandpoint, but some snow nonetheless. The mean snowfall is right around an inch, however, there is one run which shows  totals nearing 5"!

Now despite these snow forecasts, the other factor to consider is how easily will this snow accumulate on the ground? The temperature forecast for late Sunday night and early Monday morning is for readings in the lower to middle 30s. Certainly cold enough for snow, but perhaps not enough for significant accumulations, especially in the Spokane area. After sunrise, temperatures will slowly climb into the upper 30s to middle 40s which suggest snow accumulations are even less plausible except perhaps on grassy surfaces. 

So in summary we are fairly confident a deep but compact surface and upper-level low-pressure system will track through the Inland Northwest producing locally moderate to heavy precipitation some of which will fall as snow. Where it will go and what time of day it hits will be critical for determining what it's impacts will be. Winter driving conditions are certainly possible for the Monday morning commute, especially north and east of Spokane. In the meantime, stay tuned to our latest National Weather Service Forecasts and don't put away your winter clothes quite yet. 

Thursday, March 12, 2015

When does Spring really start?

We're often asked the question "when does Spring really start?"  The easy answer is to look at your calendar.  It says March 20th.  This is "Astronomical Spring", which is the vernal equinox.  It's defined as the day when the sun is directly over the equator, and the length of a day is 12 hours no matter what your latitude (this isn't quite correct, but we won't go into that here).  But it doesn't take much thought to realize that this definition is only remotely linked to the weather.  And having one date for the entire Northern Hemisphere seems in some ways, ridiculous.  Spring starts in Miami and Fairbanks on the same day?

The "Meteorological Spring" is another definition.  It's defined as the months of March, April, May.  But again, this suffers from a "one size fits all" problem.  Surely there's a better way of defining Spring that varies from one location to another.  The Capital Weather Gang at the Washington Post addressed this same topic, and had some other ideas.  You can read their blog here.  

Some of their suggestions have this natural variability built in, such as the last day of measurable snow, last freeze, soil temperature, and first green leaves.  But many of these would be difficult to apply to all locations.  What about locations that don't receive snow, never freeze, or don't lose all of their leaves?  And if a location receives snow in May, does that mean that it's still Winter?

For central Washington, the Tree Fruit Research Center in Wenatchee tracks the average first bloom of the apple trees.  You can see those dates at this link.  The nice thing about a measure that tracks leaves, flowers, or soil temperatures, is that it will vary year-to-year (i.e. spring doesn't start on the same day every year).

The intriguing measure in the Capital Weather Gang blog was defining seasons based on average temperature.  They took the average daily temperatures at a location, and divided them into the coldest 1/4th (winter) and warmest 1/4th (summer), with spring and fall as the seasons in between.  If we do this for the Spokane Airport, we get the following:

Winter: Nov 11th - Mar 6th (116 days)
Spring: Mar 7th - Jun 8th (93 days)
Summer: Jun 9th - Sep 15th (99 days)
Fall: Sep 16th - Nov 10th (57 days)

While this is somewhat interesting, we're not quite sure if we agree with its accuracy.  Fall is the shortest season, no doubt about that.  But in reality, spring is the longest of all season in the Inland Northwest.  It typically starts in late February or early March, but it often lasts all the way through June.  We have a saying that summer in the Inland Northwest doesn't start until after the 4th of July, in Seattle, they claim that their summer starts on the 12th of July.  This means that in reality, spring lasts about 4 months while summer is just a shade over 2 months long.

For us, it's often easiest to classify the seasons by holidays:

Winter: Thanksgiving to President's Day
Spring: President's Day to the 4th of July
Summer: 4th of July to Labor Day
Fall: Labor Day to Thanksgiving

Friday, March 6, 2015's been warm! Will it continue?

Over the past 30 days the Inland Northwest has been abnormally warm, and indications are the Inland Northwest will experience well above normal temperatures next week, and maybe beyond.  Before we look ahead, take a look at the temperature anomalies over the past 30 days.

Temperatures have been running about 3 to 7 degrees above normal, with the greatest anomalies over North Central Washington, with even higher anomalies over portions of eastern Oregon, and southern Idaho.

The weather pattern ahead appears to be showing a very warm signal.  So what is going to cause this?  The atmosphere often goes through various oscillations that affect our weather pattern.  One of these that we will not dive to deep into is the Arctic Oscillation (AO).  This AO value is forecast to soar next week with the highest value seen in quite some time.  Take a look at the forecast for the AO

The black line is the observed and the red line is the forecast.  As you can see, the highest value seen since November is 3.5, and is forecast to reach a value close to 6 within the next week.
This combined with our current weak El Nino often produces warm temperatures.  Take a look at what typically occurs temperature wise with this pattern:

As you can see nearly the entire US usually sees warmer than normal temperatures in this scenario.

So, let's look at the upcoming pattern.  Here is the GFS model forecast for Tuesday

The model indicates a low pressure system over the eastern Pacific Ocean with a warm southwest flow ahead of it shown by the arrow.  This will usher in warm air from the southwest into Washington, Oregon, and Idaho.

Here is our forecast high temperatures for Tuesday, which is about 15 degrees above normal and near record values for many cities.

Now, here is the GFS model forecast for Thursday

The pattern looks similar.

How about next Saturday?  Does it look any different?

The warm pattern continues.  Regarding precipitation, this flow often brings in increased moisture as well.  At this time, it looks like the heaviest precipitation through the week will remain north of the Canadian Border.  The best threat for rain for the Inland Northwest will be Wednesday into Thursday.  

What about after that, here is the 8-14 day outlook from the Climate Prediction Center:
The pattern refuses to budge!  What about after that, what is the spring looking like?  Here is the CPC outlook for March, April, and May.

It appears that this warm weather pattern may last for awhile, with elevated odds of warmer than normal temperatures.  Of course, occasional rounds of cooler and unsettled weather typically occur in the spring.  But when all is said and done, we will probably look back at this being a warm spring.

Saturday, February 28, 2015

Update on Snow Chances

In our last blog, we talked about the chances for snow in our area on Monday and/or Tuesday.  Our conclusions were:

  • Any precipitation would fall as snow.
  • The overall weather pattern was not a good one for snow in the Inland Northwest.
  • Not all of the computer guidance agreed on exactly what was going to happen.
Two days later, those conclusions still hold true.  Usually at this point we have a pretty good idea of what's going to happen.  But in this situation, we're still no more confident than we were two days ago.  Here's why.

Here's the GFS forecast for Sunday evening.  The low is way up in northern Canada with a trailing cold front (blue line).

GFS forecast of precipitation (green shading) and sea-level pressure (thin lines) for Sunday evening.

Here's the forecast for Monday afternoon.

GFS forecast of precipitation (green shading) and sea-level pressure (thin lines) for Monday afternoon.

As you can see, the front splits.  Half of it goes into the upper plains while the other half drops down the West coast.  This leaves the Inland Northwest dry, which is pretty common for this kind of storm moving down from the northwest.

Let's look at the Canadian GEM model:

Canadian forecast of precipitation (blue green shading) and sea-level pressure (thin lines) for Monday afternoon

This looks very similar to the GFS model.  The GEM does bring a bit more snow into the Panhandle, but keeps eastern Washington dry.

So why isn't our confidence higher?  Because the ECMWF has a slightly different scenario.  It thinks the cold front will hold together as it passes over our area.  So much so, that it gives about 1-2" of snow to the Panhandle and eastern Washington, east of Moses Lake.

What's more is that each model has been consistent with itself.  Often, we'll see forecasts from the computer models that will change 12 hours later.  In those situations, our confidence is lower.  But in this case, each forecast from each model is very similar to its previous forecast.  So the GEM and GFS insist that the Inland Northwest won't see much from this storm, while the ECMWF continues to insist that we will see 1-2" of snow.  

So, are there any other computer forecast models?  Yes, there are.  Here's the UKMET forecast:

UKMET forecast of precipitation (blue and green shading), sea-level pressure (thin black lines) and 500mb heights (red lines) for Monday afternoon.

You can see that the UKMET has light precipitation over all of Washington.  This is similar to the ECMWF, just lighter on the precipitation.  If this were to verify, eastern Washington might see a dusting of snow.

The SREF model is an ensemble, which means that it's actually a group of 23 similar models.  Here's a display of the SREF forecast snowfall for Spokane:

SREF forecast of snow for Spokane

Of the 23 SREF models, only 6 of them have any snow for Spokane, and only one has nearly an inch of snow.

So given all of this, the odds are slim that eastern Washington will see any snow on Monday.  Not impossible, just not likely.  Thus, the forecast of "a 40% chance of snow showers" for Spokane.  The northern Panhandle has a bit better odds, but still not a slam dunk.

After that, we don't see any significant precipitation for at least another week.