Sunday, September 30, 2018

Is Global Warming A Significant Contributor to Washington State Wildfires, Now or in the Future?

During the past several summers there have been major wildfires in Washington State producing a lot of smoke.  And many people have been asking an important question:

To what degree is anthropogenic global warming contributing to Washington State wildfires?


If 90% of the blame for Northwest wildfires is due to anthropogenic (human-caused) global warming and 10% is due to fire suppression, poor forest management, or people starting more fires, then the logical response is to put most of  our efforts into reducing atmospheric CO2.   A climate-dominated problem.

If 90% of the blame is due to past fire suppression, forest mismanagement, invasive species, and human encroachment, then we should put most of our efforts into fixing the forests and other non-climate measures. A surface-management problem.

And yes, the percentages could be somewhere in between.

Supporters of the carbon fee initiative (1631) are suggesting that the recent wildfires are mainly the result of anthropogenic climate change and using the fires to push their carbon fee plan.


And Governor Inslee has stated explicitly that the fires have been made much worse by climate change.

In contrast, others, including a number of folks in the forestry community, have suggested that poor forest practices are the main cause of most of the wildfires over the eastern side of the state. 


It is important to note that relative role of global warming in influencing the threat of wildfires may change in time.  For example, global warming could be relatively unimportant today for wildfires, but of great importance later in the century when temperatures will be much warmer.

The Need for Better Information

There is actually very little limited quantitative information on the role of global warming on Washington State wildfires.   Which is kind of strange considering the importance of the issue and the authoritative statements being made by some.  A lot of hand waving, but not much data.

So let's examine the issue in some depth, using a more quantitative approach than most.  But before I do so, let me give you the bottom line.

Human-caused global warming has played only a minor role regarding  Washington State wildfires through today, but will become much more important later in this century.

Now let me provide some evidence for this conclusion.

How Has Global Warming Changed Washington's Summer Climate?

Before we look at the correlation of global warming and wildfires, we need to know how much Washington State climate has changed during the past half century or so, and then  estimate how much of that change is due to anthropogenically forced increases in greenhouse gases. To gain some insight into this, I secured the official NOAA/NWS climate division data averaged over Washington State for summer (June through August).

First, consider daily mean temperatures from 1930 to today..  Very little warming until the mid-70s and then perhaps 2°F overall during the past 40 years.

Summer average maximum temperatures have similar  pattern of change--again roughly 2F warmer since the mid 1970s.

There is a substantial research that suggested that the radiative effects of increasing CO2 in the atmosphere became significant for climate forcing something around the 1970s.   And there was an important shift in a mode of natural variability, the Pacific Decadal Oscillation (PD0) during the mid-70s:  a shift from the cool to warm phase of this oscillation, which would have resulted in warming over the Pacific Northwest.

Now the question is how much of the recent warming shown above is due to anthropogenic global warming and how much is due to natural variability.

A group of researchers at the UW (including myself) are working on this question, using the most sophisticated approach applied to date:  an ensemble of high-resolution regional climate runs forced by the best global models.  This is the gold standard for such work.  We started with global climate models driven by the most aggressive increase of greenhouse gases (RCP 8.5) and then ran a high-resolution weather prediction model (WRF) driven by several climate models over time (1970-2100).

I don't have this output interpolated to the exact boundaries of WA state (working on this now), but let me show you the projections for summer max temperatures from the high-resolution model for two sides of the state (Hoquiam, HQM, and Spokane, GEG) forced by several climate models (see below in colors).  I also show the observed temperatures during the contemporary period at these location).  Virtually all of our simulation show greater warming at Spokane then along the coast, so let me show you that first.

Between 1970 and roughly 2000 there is very little change in observed or modeled temperatures at Spokane, and roughly1.7F warming between 2000 and now in most of the simulations.  Since natural variability will differ between the simulations, the 1.7F average of all of the runs is a reasonable estimate of the impact of global warming until now.   And note how the warming revs up later in the century if the aggressive increase in greenhouse gases continue s(about 7F warming!).

At Hoquiam, near the WA coast, the warming is less for both the recent decades and into the future.  Perhaps 1F of warming through this year.
Now, I could show you a lot more material, but my conclusions from looking at a lot of high-resolution model data is that anthropogenic global warming due to increasing greenhouse gases may well have warmed up the state as a whole by roughly 1-1.5F during the past half-century, with any additional warming coming from natural variability (e.g., the PDO).   I really doubt that there would be much disagreement about this estimate from members of the atmospheric community.

What about changes in precipitation?

Summer (June to August) precipitation has always been relatively modest (4-5 inches) in our state (our summers are very dry), and there appears to be a modest wetting trend through 1980 and some drying since the late 1990s (see below).  In contrast, annual precipitation has been very constant (also below)
Summer Precipitation for WA State (1930-2017)

WA State Annual Precipitation

What do the climate simulations suggest about precipitation trends?

 The annual precipitation will slowly increase according to the models (not shown), but what about summer?  Spokane summer precipitation has always been low (around 2.5 inches typical for June through August) and will remain low.  Any decline is small--a half-inch at most.  The recent dry years look like natural variability.

Let's compare that to Seattle on the western side of the State.   Summers are equally dry as Spokane, but there is a more clearcut drying-- by roughly 2 inches by 2100, and perhaps .5 inches during the past years.  These results are consistent with previous studies by the UW climate impacts group and others indicating a slow increase of total annual precipitation, but a small downward trend in summer precipitation over our area.


So to summarize.   Looking at past climate data and the best model information, one concludes that increasing greenhouse gases may well have warmed out state by 1-1.5F during summer (June through August) over the past half century, had little impact on annual precipitation, and perhaps dried an already very dry summer by perhaps .5 inches.

But how did global warming impact the recent wildfires in Washington State? And how will future warming impact them?

We are now ready to answer this question.

But first we needed a list of the annual area burned and number of fires in Washington State over time.  It turns out this is a difficult information to get--which is surprising considering its importance.  I was able to get an Excel file from Josh Clark of WA State DNR with the fire information from 1992 to the present.  The prior period has not been digitized, with fire information in cabinets somewhere.  Oregon and California has done a better job in creating a long-term digital record of their fires.

OK, we will use what we have.  Here is the number of acres burned by wildfires over WA state since 1992.  A very slow trend upward, except for the HUGE peak in 2015, the year with the big ridge and crazy-warm spring.



The number of fires (see below) have been nearly constant in the long term, with some ups and downs/


But now the really interesting part.  Let's plot the acres burned against warm season (June through September) temperatures (see below).

This is really fascinating.  A very slow increase of fire acres with temperature, with considerable scatter,  showing that acres burned is not that temperature sensitive.  The one big year was the warmest.. 2015 with 61.4 F and nearly 1.2 million acres burned.



Now, let's put a regression line on this plot and see how much of the variability is explained by increasing temperature. Temperature only explains 22% of the increase in acres burned...so 78% is explained by something else.
The bottom line of all this is that warming temperatures can explain only a small portion of the variability in Washington wildfires.

What about precipitation?   Here is the plot of WA state precipitation for May through Sept since 1992.  A very slight downward trend, with the big fire year (2015) not showing anything anomalous.

Another "scatterplot", this time of acres burned versus precipitation, is presented below.  Very poor relationship, with the suggestion of a decline in burned acreage with greater rainfall.  And the precipitation only explains about 2% of the variability of acres burned!  This is not surprising because our region is naturally dry during the summer and being a little drier doesn't make that much of a difference.   Like being a little more dead.


The Essential Message Here

  Climate/weather changes do affect wildfires over Washington State.  Warmer temperatures and lesser precipitation correlate with increasing acreage, but the relationship is not a strong one.  The correlation of summer precipitation with wildfire acreage is very, very weak and summer temperature only explains less than a quarter of the variability in wildfire area.   

Then we look at the look at the amount of climate change produced by human-caused greenhouse warming so far, and we find it is relatively small. Perhaps 1.5F for WA State temperatures and a slight drying over the summer.

You put the lack of sensitivity together to temperature/precipitation with the small climate changes due to global warming and one has to conclude that human-caused climate change is undoubtedly NOT a major driver of the increased wildfires and wildfire smoke we have seen during some recent years.

Based on my extensive reading on the wildfire issue, discussions with forestry experts at the UW, and a number of seminars/meetings I have attended, my conclusion is that the real culprits for our invigorated fire/smoke situation include:

1.   Nearly a century of fire suppression and poor forest management that have produced unnatural, explosive forests, particularly on the dry side of our state.
2.  Huge influx of people into the urban-wildland interface and forest areas that help initiate fires and make us more vulnerable to them.
3.  Invasion of highly flammable, non-native species like cheatgrass.


And we should not forget that fire is a natural part of our east-side forests.

Claiming the climate change is the big villain in the current wildfire situation, may be a useful tool for some ambitious politicians and for those searching for arguments to support climate-related initiatives, but the truth is probably elsewhere.  In the FUTURE, as temperatures warm profoundly (particularly during the second half of the century), the influence of human-produced global warming on our wildfires will clearly increase substantially.

Only by a sober, fact-driven approach, such as thinning, debris-removal, and proscribed burning of our east-side forests, with will be able to improve the health of our forests and reduce the potential for megafires and big smoke production.  Even if we could stop anthropogenic climate change in its tracks this year, we still need to  deal with the issues of forest management, human initiation of fires, and human changes at the surface.

PS:  Although we had considerable background smoke from Canada, the really extreme smoke periods (August 21-22 in Puget Sound) was associated with fires over NE Washington, not Canada.  Same thing in 2017, with WA fires resulting in ash falling on Seattle.

PSS:  Some folks might bring up the Pine Beetle issue.  I have read several papers and talked to experts in UW Forestry that suggest that rather than lack of cold temperatures, unnaturally dense east-side forests and lack of fire allowed the Beetle kill.  In any case, peer-reviewed papers suggest that pine beetle infestation does NOT contribute to fires.

____________________________________
A local forest landowner named Michael August has written a very interesting perspective on NW forests and smoke, found here.

Friday, September 28, 2018

Fall: The Time of Big Temperature Swings

We are now going into the season of very large temperature swings between the daily maxima and minima.  The period between the end of summer and the beginning of the clouds and rain.

Days are sunny and  the sun still has some strength, resulting in decent daytime warming.

Nights are much longer and with relatively clear skies, there is good radiational cooling to space.


To illustrate, here are the high and low temperatures today (click on maps to enlarge).  In western Washington we were in the mid to upper 70s....warm for this time of the year.  But western Oregon was in the mid 80s to 90s, while lower to mid-90s were observed over northern CA. Mid-80s over the high desert of southeastern Oregon.


But now look at the minima.  Wow.   Mid-40s in western Oregon and Washington. And 30s and even below freezing temperatures in eastern Oregon.

Can you imagine, going from mid-80s to around freezing in a matter of a few hours?   That would crack some concrete!

Need more convincing?  Here are the plots of temperature at Olympia,WA and Medford, Oregon (central eastern OR) for the past three days. Olympia (purple line) ranged from the mid-40s to the mid-70s today, but Medford went from 50 to roughly 90F. 

But even these swings are eclipsed by a few stations in Oregon and northern CA, where yesterday some went from 100F to the low 40s.

So if you don't like the weather, wait a few hours.  My advice:  get outside tomorrow (Saturday)....clouds and light showers expected on Sunday.



Wednesday, September 26, 2018

Winter Forecast for the Northwest

I have been under a lot of pressure to provide the winter forecast.  And the snow outlook is probably the biggest demand. 

OK, you will get what you want.
What some folks want

The skill of our forecasting models really fade after two weeks with the extended prediction models having virtually no skill months ahead of time.  The most useful tool we do have is the correlation of the temperatures of the tropical Pacific, quantified as El Nino (warm water), La  Nina (cool water) or neutral conditions (La Nada), and the weather over the Northwest.     The correlation is not perfect, but gives one an idea which way the atmospheric dice are weighted.

Let's start by look at the temperatures of the tropical Pacific, with the temperature of the Nino 3.4 area being the most closely monitored (see map).


The sea surface temperature (SST) anomaly (difference from normal) is shown below.   As you can see, the Nino 3.4 area  this spring went from colder than normal (La Nina) to weakly above normal.    Neutral or La Nada conditions area associated with the SST anomalies being between .5C and .5 C.    That is where we are now.

But it appears, we will get to at least to a weak El Nino (Nino3.4 temperature anomalies between +.5 and 1C).    One hint is that the temperature of the upper ocean is warming (see below)
And if we look at a collection of forecast models, they are pushing the temperature anomaly to about 1C above normal in the Nino 3.4 area (see below).  A weak El Nino.

 The official Climate Prediction Center prediction is for us to move from neutral to weak El Nino, with roughly a probability of 65%.


But we have an issue.  To get a significant El Nino signal, one that really changes the midlatitude weather pattern, one needs a moderate to strong El Nino.  In those circumstances, the Northwest is generally warmer than normal, the jet stream splits over the northeast Pacific and good portion of the jet stream heads south towards southern CA.


The El Nino this year appears to be a marginal one, with far more limited impacts.   Even in a weakened state, the El Nino tendency would be for warmer than normal conditions and less snow than normal.   And add to that the decadal trend to slowly increasing temperatures from global warming, and the fact that an important source of natural decadal variability, the Pacific Decadal Oscillation (PDO), recently moved into a warm phase. (see below)


So warmer than normal seems a good bet. 

Let's take a look at one of the best extended forecasting modeling systems (but still with marginal skill), the North American Multi-Model Ensemble (NMME).  I will show you the average anomalies (differences from normal) for precipitation and temperature for November through January.

Precipitation?  Drier than normal over much of the Northwest, but close to normal over western Washington.
 Temperatures?   Around .5C above normal.

The implication of all this would be a lower than normal snowpack in our mountains this winter.  Lower heating bills.  A slow fill for our regional reservoirs.   

One thing I do know...the next three days will represent the most perfect fall weather you can imagine.  Perfect day time temps, cool nights, leaves starting to turn.  My favorite season.

Monday, September 24, 2018

Super Convergence Zone Dumps Heavy Rain

One of the most striking peculiarities of western Washington weather is the Puget Sound convergence zone  (PSCZ), which can produce a heavy band of precipitation across northern Puget Sound and into the Cascades.

The PSCZ  results when low-level winds on the coast are westerly (form the west), move around the Olympics, and then converge together over Puget Sound, producing upward motion, clouds and precipitation.

This is exactly what occurred Saturday evening, as illustrated by the plot of surface observations at 7 PM (a large scale and close up view is shown below)



The air was relatively unstable Saturday evening and an upper level trough was producing additional uplift....so the convergence zone really revved up.    

Now let me impress you...here is a radar image around 7 PM Saturday.  Red indicates very heavy rain, and yellow is moderate to heavy. Intense convergence zone band.  Note the sharp southern edge of the rainfall, which is very typical of strong convergence zones.


There was substantial convection and thunderstorms in that heavy precipitation band, and lots of lightning was noted by the regional lightning sensors (see below-24h lightning ending 1 AM Sunday)



Seattle RainWatch combines radar and observations to provide a good estimate for rainfall.   Here is the RainWatch 24-h total ending 5 AM Sunday.  A substantial area got 1-2 inches, which even more in limited regions.  Pouring over Snohomish County but downtown Seattle was dry.   You got to love living around here.

Local rain gauges had 24-h rainfall totals that were consistent with RainWatch, with over 2 inches around Monroe.


I know someone that had a wedding celebration near Monroe...thankfully they had tents, but it was a soggy affair.

Now the really exciting part for me is how well this event was forecast....our high-resolution models was very skillful, days ahead of time.  To illustrate, here is the forecast from the UW WRF high-resolution (4/3 km grid spacing) model for the 24h precipitation ending 5 AM Sunday.  

Nice convergence zone.  It underplayed the total amounts a bit...but it was clear that Snohomish County was going to be drenched, which downtown Seattle would be dry.


Saturday, September 22, 2018

REX Will Dominate Northwest Weather This Week

During the past two weeks, the Northwest has enjoyed absolutely normal weather.   Typical temperatures (see plot below, purple and cyan are normal highs and lows) and nearly normal precipitation (see cumulative precipitation for the same period).
 Normal has been good for us, giving relief to our plants, restoring water to the surface soil layers, and radically reducing the water usage in urban areas such as Seattle (see below).


The occasional precipitation has been associated with transient upper-level troughs that have moved through the region, such as one moving across the Northwest this morning (see upper level, 500-hPa, ap for 5 AM this AM)


But this week something interesting is going to happen:  a very stable REX BLOCK will develop over the eastern Pacific, shutting off precipitation, leaving us with lots of sun, dry conditions, and moderate temperatures. 

It will be wonderful.

So what is a Rex Block?   It is a configuration of the atmosphere where the upper level circulation has a ridge of high pressure (or heights) north of a trough of lower pressure/heights.  Here is a schematic of such a situation.


A Rex Block is a very stable atmospheric configuration, with the ridge and trough reinforcing each other, keeping the flow pattern in place  (for reasons I talk about in my graduate synoptic/dynamics class, but won't go into here).

The ridge/high part produces descending motion and fine weather, particularly on its eastern side.  Storms are sent far northward--in the example shown, well into Alaska where they belong.

The Rex Block has nothing to do with kings or dinosaurs, but with a meteorologist named D. F. Rex, who wrote a seminal paper on this features way back in 1950 (see below).

Rex, D. F. (1950). "Blocking Action in the Middle Troposphere and its Effect upon Regional Climate". Tellus. 2 (4): 275–301.

Now that you are Rex-trained, lets look at the upper-level forecasts for this week.

11 PM on Sunday?  Rex is here!  Huge ridge of high pressure/heights over the northeast Pacific, with a low underneath.


 2 PM on Tuesday....classic Rex!


 5 AM on Thursday.   A Rex fiesta.


The REX block shifts westward on Friday (see below), allowing a trough to move southward over us late Friday night.   Some showers and cooling if it happens.


Will the REX block reestablish itself?   Stay tuned.