July 2016 Observation and Model Discussion for the Pacific Northwest

[Phil]

...........thanks?

Lol, that was random even by Dewey standards.

[Phil]

I think either the SEA sensor needs a new/better radiation shield, or there's contamination. Jumped 6 degrees in 10mins, then dropped 4 degrees in 5mins, all without significant variations in wind/cloud cover.

 

If sunbreaks are causing 5+ degree swings in temperature, that's a problem with the sensor/shield. I have three temperature sensors here (SHT11, SHT13, SHT15), and when swings like that start showing up/timed w/ sunbreaks, I replace them and the problem is solved.

[Phil]

The sensor @ DCA was replaced last summer after similar behaviors were noted, and the problem is largely fixed to this day.

 

I'm just mentioning this because, to warm a typical near-surface thermodynamic environment 6 degrees in 10 minutes, assuming typical emissivity, convective transport, latent heat conversion/EC rates, and thermal capacity of the conducting surface, would require an an extraordinarily large amount of energy, more than could be accounted for via sun breaks (especially given SEA's latitude/solar vector).

[Phil]

So far, the 12z EPS is cooler than average for SEA/PDX through day 11. It wouldn't surprise me if (in reality) a few above normal days occur during the retrogression process, but the long term signal is a fairly robust one.

[TT-SEA]

Neither Friday or Sunday will be above normal. SEA will probably be running around -3.5 or so by 7/15, give or take a degree.

 

 

Today is going to be normal or above normal at SEA... Sunday is likely candidate as well.

 

GFS MOS and WRF served me well.  

 

69 at SEA and at UW now.  

 

Should end up around 73 or 74 after a low of 58.

[Deweydog]

Lol, that was random even by Dewey standards.

 

I give his post a -10.  

[Phil]

Today is going to be normal or above normal at SEA... Sunday is likely candidate as well.

 

GFS MOS and WRF served me well.

 

69 at SEA and at UW now.

 

Should end up around 73 or 74 after a low of 58.

I believe the average 24hr temperature determines the daily anomaly. So, SEA will need to average at least 66 degrees to obtain a positive anomaly.

 

As of now, SEA is at ~63.5 degrees since midnight.

[Jesse]

I believe the average 24hr temperature determines the daily anomaly. So, SEA will need to average at least 66 degrees to obtain a positive anomaly.

 

As of now, SEA is at ~63.5 degrees since midnight.

This isn't true. It's just the average of the high and low.

[James Jones]

The sensor @ DCA was replaced last summer after similar behaviors were noted, and the problem is largely fixed to this day.

 

I'm just mentioning this because, to warm a typical near-surface thermodynamic environment 6 degrees in 10 minutes, assuming typical emissivity, convective transport, latent heat conversion/EC rates, and thermal capacity of the conducting surface, would require an an extraordinarily large amount of energy, more than could be accounted for via sun breaks (especially given SEA's latitude/solar vector).

My guess is those readings are just bullshit, nothing wrong with the sensor. PDX has reported some whacky 5 minute swings too, and there are often reported readings that are higher/lower than the day's actual high/low.

[Phil]

This isn't true. It's just the average of the high and low.

That's a stupid way to determine a daily average. Wow.

 

That's like the score of a baseball game being determined just by the runs scored in the 1st and 9th inning. You're literally excluding the vast majority of the data when merely averaging the high/low.

[Phil]

My guess is those readings are just bullshit, nothing wrong with the sensor. PDX has reported some whacky 5 minute swings too, and there are often reported readings that are higher/lower than the day's actual high/low.

Well, BS readings won't occur if the sensor and radiation shield(s) are functioning properly.

[Timmy Supercell]

Even though I had quite a period of moderate rain (even a few minutes of heavy rain) KLMT manages only 0.01" of rain. lol

 

Nice day today! Off and on light rain and in between showers is a nice humid 62 degrees. Feels like I'm back in Hillsboro in mid May

[Phil]

It's true, though, I think. If you have a 12:01 a.m. high of 70 and then the temperature drops to 40 within the next hour and stays there the rest of the day, I believe the day's average temperature still goes in the books as 55.

 

Not that that would ever happen around here, though.

Haha, that's mathematical incompetence of the highest possible degree. Even elementary school kids know how to properly average a set of numbers.

[Jesse]

Haha, that's mathematical incompetence of the highest possible degree. Even elementary school kids know how to properly average a set of numbers.

Don't they do it that way in your area too?

[Deweydog]

That's a stupid way to determine a daily average. Wow.

 

That's like the score of a baseball game being determined just by the runs scored in the 1st and 9th inning. You're literally excluding the vast majority of the data when merely averaging the high/low.

Are you just now realizing this is how daily averages are determined?

[Abbotsford_wx]

FWIW, my weather software calculates the average both ways, and I find there's very little difference in the end.

 

Averaging the minutely data, my average for July so far is 61.3F

Averaging just highs and lows, it's 61.7F

[BLI snowman]

FWIW, my weather software calculates the average both ways, and I find there's very little difference in the end.

 

Averaging the minutely data, my average for July so far is 61.3F

Averaging just highs and lows, it's 61.7F

 

We rarely get sharp fronts here, especially this time of year.

 

You'll notice a dramatic difference if you're talking about Denver or Billings.

[Phil]

Don't they do it that way in your area too?

I guess they do. I really haven't looked into how they compute averages, honestly. I just assumed it was done competently (apparently I was wrong).

[Phil]

Are you just now realizing this is how daily averages are determined?

Yeah, pretty much.

[Abbotsford_wx]

We rarely get sharp fronts here, especially this time of year.

 

You'll notice a dramatic difference if you're talking about Denver or Billings.

 

On a daily basis, yeah. I meant more over the course of a month... it tends to smooth out to within a few decimal places. Last July for example, I get identical values calculating both ways. Most months I've checked are within 0.4F.

[Deweydog]

Yeah, pretty much.

Surprising.

[Phil]

Surprising.

Lol, yeah who woulda thunk it?

 

I just assumed that daily averages were actually daily averages, rather than shoddy interpolative splices. Just when I thought climatological data standards couldn't get any lower..

[Deweydog]

Lol, yeah who woulda thunk it?

 

I just assumed that daily averages were actually daily averages, rather than shoddy interpolative splices. Just when I thought climatological data standards couldn't get any lower..

Seems a little overly dramatic. Even in areas which can see dramatic changes in air mass over a short period of time, car more often than not an average of the high/low is gonna be a pretty fair representation.

[Phil]

Seems a little overly dramatic. Even in areas which can see dramatic changes in air mass over a short period of time, car more often than not an average of the high/low is gonna be a pretty fair representation.

"Pretty fair" is a pretty low standard. It's easy to compute a 24hr average, so this is laziness more than anything else.

[Deweydog]

"Pretty fair" is a pretty low standard. It's easy to compute a 24hr average, so this is laziness more than anything else.

Outrage.

 

Anomaly Lives Matter!!!

[Jesse]

"Pretty fair" is a pretty low standard. It's easy to compute a 24hr average, so this is laziness more than anything else.

 

You make a good point here.

[Phil]

Outrage.

 

Anomaly Lives Matter!!!

Lawyers make terrible scientists.

[Deweydog]

Lawyers make terrible scientists.

Good thing I'm not a scientist.

[Phil]

Good thing I'm not a scientist.

Clearly.

[Deweydog]

Clearly.

I have too much common sense for my own good.

 

Do you think they should change? Is it not the consistency and accuracy of readings that is most important when looking at a large collection of data?

[Jesse]

I have too much common sense for my own good.

 

Do you think they should change? Is it not the consistency and accuracy of readings that is most important when looking at a large collection of data?

 

Wow.

[Jesse]

Some showers starting to fire up.

[Deweydog]

Wow.

 

Case in point.  You'd think someone who intensely studies climate would know how daily temperature averages are calculated.  Common sense.  

[Phil]

I have too much common sense for my own good.

 

Do you think they should change? Is it not the consistency and accuracy of readings that is most important when looking at a large collection of data?

Lawyers also think they're smarter than they actually are because they know how string words together.

 

I guess it comes down to whether you want the *real* average or not. We're not talking long term stuff here, however there are also many climates (think sea breeze/monsoon climates) that will be poorly represented via this sort of interpolative averaging.

[Phil]

Case in point. You'd think someone who intensely studies climate would know how daily temperature averages are calculated. Common sense.

Common sense would be computing averages properly.

[Jesse]

Case in point.  You'd think someone who intensely studies climate would know how daily temperature averages are calculated.  Common sense.  

 

I get what you are saying, just coming off a little full of yourself, and also implying scientists somehow lack common sense, which is sort of an unfair (and inaccurate) stereotype.

 

I can see the merit of doing a 24-hour temperature average for accuracy's sake, but then again the fact that we have been doing it this way for so long holds some weight. If we changed it now it would be really hard to compare pre-24 hour average records to post.

 

Also, in the long term, you will probably get fairly equal representation of hourly reading outlier events that cut both ways. A 32 at 1am before a warm front moves through and puts the rest of the day in the 50s would be just as "misrepresented" by simply averaging the highs and lows as a day that is in the 70s all day and plunges into the 40s a few hours before midnight. In both cases, that method doesn't really paint an accurate representation of the majority of either day, but in the end I would think it shakes out with the two types of days more or less cancelling each other.

[stuffradio]

18Z is back to showing warmth and is more dry within 10 days. It also has temps in the low 80's at the end of next week.

 

They are going to do a news segment about our colder/rainier weather too tonight.

[Phil]

18Z is back to showing warmth and is more dry within 10 days. It also has temps in the low 80's at the end of next week.

 

They are going to do a news segment about our colder/rainier weather too tonight.

Looks like 13/16 days are cooler than average, with another round of heavy troughing following the ridge retrogression.

[Deweydog]

I get what you are saying, just coming off a little full of yourself, and also implying scientists somehow lack common sense, which is sort of an unfair (and inaccurate) stereotype.

 

I can see the merit of doing a 24- hour temperature average for accuracy's sake, but then again the fact that we have been doing it this way for so long holds some weight. If we changed it now it would be really hard to compare pre-24 hour average records to post.

 

Also, in the long term, you will probably get fairly equal representation of hourly reading outlier events that cut both ways. A 32 at 1am before a warm front moves through and puts the rest of the day in the 50s would be just as "misrepresented" by simply averaging the highs and lows as a day that is in the 70s all day and plunges into the 40s a few hours before midnight. In both cases, that method doesn't really paint an accurate representation of the majority of either day, but in the end I would think it shakes out with the two types of days more or less cancelling each other.

 

That's my point.  Changing now would only diminish the scientific weight of what has been a consistent, accurate method.  

 

Another aspect to consider if you look at it from the Phil-angst approach is in the old days the data collection would have been exhaustive.  Applying today's immediate information gratification standard is not reasonable.  If someone wanted to go back now and recalculate it for every official reporting station across the nation or whatever they would most certainly uncover some VERY disturbing disparities.  Or not...

[Phil]

That's my point. Changing now would only diminish the scientific weight of what has been a consistent, accurate method.

 

Another aspect to consider if you look at it from the Phil-angst approach is in the old days the data collection would have been exhaustive. Applying today's immediate information gratification standard is not reasonable. If someone wanted to go back now and recalculate it for every official reporting station across the nation or whatever they would most certainly uncover some VERY disturbing disparities. Or not...

I completely disagree with this, but considering I've been pissed and miserable all day (broken AC with 80 degree dewpoints) I will digress for the time being.