Central Pa. Summer 2024

[Bubbler86]

The 1980's had 3 winters with no days 21 and below in Jan. 

 

[Itstrainingtime]

1 minute ago, Mount Joy Snowman said:

Windy as all get-out here in Harrisburg with cells popping all around but looks like my house and Maytown will be missed once again. 

 

[Bubbler86]

Just now, Bubbler86 said:

The 1980's had 3 winters with no days 21 and below in Jan. 

 

Mike and I had a bet as to how many posts it would take before MJS would take the train home early to get into the fray. 

[Itstrainingtime]

Just now, Bubbler86 said:

The 1980's had 3 winters with no days 21 and below in Jan. 

 

The 1980s were awful - thank god for the blizzard of 1983 because outside of that...it was torture watching cirrus above while the southern Piedmont got smoked. 

[Itstrainingtime]

MU's weather center has "only" gotten up to 98 today. 

[WmsptWx]

The last 20+ posts...

[Mount Joy Snowman]

2 minutes ago, Bubbler86 said:

Mike and I had a bet as to how many posts it would take before MJS would take the train home early to get into the fray. 

These days there is no getting home early.  Amtrak is doing track work between Lancaster and Harrisburg, and as a result, all midday trains are canceled.  They provide bus service instead, which absolutely no one utilizes haha.  4:25 is the earliest train I can take home.  Normal full service is slated to return in November, perhaps earlier, as I've heard they are ahead of schedule.  Amtrak being ahead of schedule on anything would be a first for them ha. 

[Bubbler86]

@Jns2183 I get that I am not comparing apples to apples with Jan and July.   On earth we do not have a plane of possible temps that can extend out to any given number.  During the winter we have much more room to go AN OR BN than we have room to go AN in the summer.  But the ratio of how often it gets a certain number AN and BN should not be as skewed IMO.   I have no issue with people saying 100 is a fantastical number because it is rare and is something to talk about.   

One more stat:  July days it has hit 90 and above vs. days it has stayed 84 and below from 2010 up.  It is now more common to hit 90 and above than it is to stay 84 and below.

 

 

 

 

[Bubbler86]

5 minutes ago, Mount Joy Snowman said:

These days there is no getting home early.  Amtrak is doing track work between Lancaster and Harrisburg, and as a result, all midday trains are canceled.  They provide bus service instead, which absolutely no one utilizes haha.  4:25 is the earliest train I can take home.  Normal full service is slated to return in November, perhaps earlier, as I've heard they are ahead of schedule.  Amtrak being ahead of schedule on anything would be a first for them ha. 

I started reading this and thought you were going to say the tracks melted today. 

[Itstrainingtime]

2 minutes ago, Bubbler86 said:

I started reading this and thought you were going to say the tracks melted today. 

I noticed yesterday that Amtrak was reducing speeds on the NE corridor due to the extreme heat. 

[Bubbler86]

Just now, Itstrainingtime said:

I noticed yesterday that Amtrak was reducing speeds on the NE corridor due to the extreme heat. 

MDT has to shut down at 120. 

[Jns2183]

JNS, it takes one to know one, so tell me, where did you study your Statistics?  Mine was at Penn state, where I also tutored Stats courses, but that was many moons ago and I'm losing my touch with every passing year.  I've been impressed with your statistical knowledge displayed on this board. 

Haha. I went to Penn State, finished up at Penn State Harrisburg. I spent 2 1/2 years in Engineering before switching to Accounting. When I graduated I was one math course short of a minor, but took several upper level stat courses. I went back to bartending because I loved it and I could after my brother sold a business I helped him start. Before that I worked for government as corporate tax auditor dealing with complex financial instruments, tax shelters, etc.
I enjoyed the intellectual aspect of my work but the bureaucracy really suck the living soul out of me and made me hate getting up every morning so I jumped at the chance of being able to not do that anymore. Are you statistics in figuring out what to even look for, and it's worth looking at, and in the understanding of the financial instruments themselves. Although numbers took a back seat to trying to figure out the legal framework and how to apply different things.

Feel like I know enough about statistics to get myself in trouble and make myself dangerous by deluding myself. It's very difficult trying to put some of these concepts in simple terms and actually I have found out that some of the best teachers out there have been from Khan academy and some other amazing YouTubers. The biggest hurtle I've had was my resistance to learning python and working with big data and although I'm still just an amateur with it I'm slowly getting better. Chat GPT has been an absolutely amazing resource for learning how to apply Python to these data sets

Sent from my SM-G970U using Tapatalk

[Jns2183]

But, a lot of people do not care about that.  These numbers are what Joe Weather watcher cares about (non-Met people).    I am not trying to say the average at MDT is wrong vs. explain what the everyday person sees and feels regardless of how often.  100% agreed that the 13AN in July is much less common than the one in Jan but so is the 13 BN day...the ratio is just drastically different. 

See I think the average Joe is going to care much more about a 13 below normal day in January or a 13 above normal day in July then they will for a 13AN in January or 13 BN in July. The numbers don't matter it's how it feels to them. I remember you were asking before why new stations don't make such a big deal out of days with a much bigger above normal spread than we currently have when it's winter. And I see the answer that is because it feels good in winter versus it feeling not good in summer with a much smaller AN

Sent from my SM-G970U using Tapatalk

[Jns2183]

[mention=4667]Jns2183[/mention] I get that I am not comparing apples to apples with Jan and July.   On earth we do not have a plane of possible temps that can extend out to any given number.  During the winter we have much more room to go AN OR BN than we have room to go AN in the summer.  But the ratio of how often it gets a certain number AN and BN should not be as skewed IMO.   I have no issue with people saying 100 is a fantastical number because it is rare and is something to talk about.   
One more stat:  July days it has hit 90 and above vs. days it has stayed 84 and below from 2010 up.  It is now more common to hit 90 and above than it is to stay 84 and below.
 

 
 
 

But it has to be skewed due to the whole distribution being skewed

Sent from my SM-G970U using Tapatalk

[Bubbler86]

2 minutes ago, Jns2183 said:

But it has to be skewed due to the whole distribution being skewed

Sent from my SM-G970U using Tapatalk
 

Some skew but to go from approx. 60% more to 400% more is a jump.  Maybe it is normal when you do your magic, but I still wanted to see the plain numbers out and in sight.    I minimized the dataset to show that it is now more normal to go 3AN in July than 3BN.  

[Jns2183]

For it not to be skewed, at minimum, The median would have to equal the mean but that's not reality.

Sent from my SM-G970U using Tapatalk

[Jns2183]

Some skew but to go from approx. 60% more to 400% more is a jump.  Maybe it is normal when you do your magic, but I still wanted to see the plain numbers out and in sight.    I minimized the dataset to show that it is now more normal to go 3AN in July than 3BN.  

What are you talking about in reference to 60% more to 400% more?

Sent from my SM-G970U using Tapatalk

[Bubbler86]

1 minute ago, Jns2183 said:

For it not to be skewed, at minimum, The median would have to equal the mean but that's not reality.

Sent from my SM-G970U using Tapatalk
 

Very true.  I feel like the point of my original post did not come across well.  Stated in English, it used to be more normal for July days of 13 or more degrees BN than it is now.   

[Bubbler86]

19 minutes ago, Jns2183 said:

What are you talking about in reference to 60% more to 400% more?

Sent from my SM-G970U using Tapatalk
 

It is actually closer to 300% not 400% but still high in my opinion.    In all of MDT's July historical readings it has gone 13 or higher AN 44 times and 13 or below normal 125 times or close to 300% more likely to be 13BN vs 13AN.   I understand the 13AN normal is bumping into our statistical possible max high but 300% is still a large difference.   In January the same ratio between how often it got 13AN vs 13BN drops to 60-70%.    I am using our current 30 year norm reading to do those numbers so trying to do them with all norm readings would require your python. 

[Bubbler86]

MDT is down to 91.  Their 5 min readings say they hit 100 but it was only 2 brief readings.  Sometimes those round down....will be interesting to see if they got it.   I think they probably did. 

[Jns2183]

Very true.  I feel like the point of my original post did not come across well.  Stated in English, it used to be more normal for July days of 13 or more degrees BN than it is now.   

That's a great question and you are probably correct. The data set I have downloaded is KMDT from 1/1/1941 through 12/31/2023. The question I have is do you want to know what the probability of 13 below normal days are for a rolling 20 year. Referencing the mean during each rolling 20-year period or do you want to reference the mean for the entire data set. Basically if the average July mean for entire set is 86, and the mean for July 1950 - July 1970 is 84, the 13 BN will be different temps

Sent from my SM-G970U using Tapatalk

[Jns2183]

It is actually closer to 300% not 400% but still high in my opinion.    In all of MDT's July historical readings it has gone 13 or higher AN 44 times and 13 or below normal 125 times or close to 300% more likely to be 13BN vs 13AN.   I understand the 13AN normal is bumping into our statistical possible max high but 300% is still a large difference.   In January the same ratio between how often it got 13AN vs 13BN drops to 60-70%.    I am using our current 30 year norm reading to do those numbers so trying to do them with all norm readings would require your python. 

I think this has to be related to the theoretical max temp versus our theoretical min temp. Basically the closest you have to each extreme the more rare it is. But also seems like you're working with different statistical distributions for the high end versus the low end

Sent from my SM-G970U using Tapatalk

[Jns2183]

I'm made we got skipped over by storms once more

Sent from my SM-G970U using Tapatalk

[Bubbler86]

1 minute ago, Jns2183 said:

That's a great question and you are probably correct. The data set I have downloaded is KMDT from 1/1/1941 through 12/31/2023. The question I have is do you want to know what the probability of 13 below normal days are for a rolling 20 year. Referencing the mean during each rolling 20-year period or do you want to reference the mean for the entire data set. Basically if the average July mean for entire set is 86, and the mean for July 1950 - July 1970 is 84, the 13 BN will be different temps

Sent from my SM-G970U using Tapatalk
 

That was what I referenced above (I called it normal, you call it mean.).  The mean changed and I was using the mean for this period we are in.  Personally, there are all round numbers and it matters not what the mean was 50 years ago.  In the entirety of MDT's history, it got 13 BN almost 300% more often than it got 13AN using today's means.    Unfortunately, you are missing 50-60 years to compare to my numbers so I would have to change to 1941....since 1941 MDT has recorded 100/13An today's mean 29 times and 74/13BN 95 times so it is not far off the ratio for the entire data set.   Close to 300%. 

[Bubbler86]

9 minutes ago, Jns2183 said:

I think this has to be related to the theoretical max temp versus our theoretical min temp. Basically the closest you have to each extreme the more rare it is. But also seems like you're working with different statistical distributions for the high end versus the low end

Sent from my SM-G970U using Tapatalk
 

Agreed, but the 300% number still seems high to me.  It is obviously going to be more common to go down into a number that is statistically possible.  I feel the 300%/amount of instances is much more common before the last 25-50 years though.  Which was my original point.  Although it is still near 300% since 2000 the times it has happened on either side are by ratio much lower.    Highs in the 90's are more common now than in the past and highs in the mid 70's are less common. 

[CarlislePaWx]

High temp exactly 100.0 degrees about 45 minutes ago (4:15), then some clouds came in and sent the temp back down to 98.6.  The sun's been back for awhile but temp is very slow to respond.  Only back up to 99.0.  I'm leaning towards not exceeding 100.0 again today.  I think it's been a few years at least since I last recorded 100.  Before today I had reached 98-99 on at least 4 or 5 days so far this summer.  Sun's still out but temp drop to 98.2.

[Bubbler86]

Of note for July.  Percentages below assume once per year just for display purposes of a ratio showing how often during that period.  

MDT hit 100 or above 39 times between 1888 and 1999....39/111 is 35%.

They have hit 100 5 times since 2000 or 5/25=20% (not counting they may have today which would make it 24%)

 

 

MDT stayed at 74 or below (between 1888 and 1999) ~ 175 times or 175/111 about 150%.

MDT stayed at 74 or below (between 2000 and 2024) 16 times or 16/25 about 64%.

 

 

Both percentages on ratio since 2000 are lower.  Smaller dataset but 25 years is nothing to sneeze at.   The 150 vs. 64 is drastic.   A 234 percent difference.   MDT has not had a July day stay 74 or below since 2018!

 

 

 

 

 

 

[paweather]

1 hour ago, Itstrainingtime said:

The 1980s were awful - thank god for the blizzard of 1983 because outside of that...it was torture watching cirrus above while the southern Piedmont got smoked. 

And that’s the thing we have always had ups and down when it comes to temps in the winter months. There is no trend.

[Jns2183]

And that’s the thing we have always had ups and down when it comes to temps in the winter months. There is no trend.

There definitely is a longer term trend, but it is less drastic than people make it out to be. It's also unclear if and how much it may be related to longer term cycles. People don't handle year to year variability well.

Sent from my SM-G970U using Tapatalk

[Bubbler86]

Just now, Jns2183 said:

There definitely is a longer term trend, but it is less drastic than people make it out to be. It's also unclear if and how much it may be related to longer term cycles. People don't handle year to year variability well.

Sent from my SM-G970U using Tapatalk
 

The monster which was July 2020 seems forgotten already.