Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
On the topic of heat island, I think it might surprise some on this thread that for my area of expertise of the Southeast US, urban stations are actually closest to what the natural climate of the area would be like - at least in terms of low temperatures.
Some of you may have noticed airport stations in the Southeast US often have shockingly large diurnal ranges and surprisingly short frost free seasons, quite the contrary to what would be expected of a low elevation, humid climate. This is actually rather artificial: airport stations are cleared of the natural pine forest cover and so cool off far, far more at night than they would if they were forested due to enhanced sky exposure.
It's not like the urban heat island would produce overreading either: most Southeast US cities are not very urban at all despite those cold airports so if anything urban heat island may underread compared to a properly forested station for places like say Valdosta. I've mentioned all the above plenty in the first/last frost threads but hadn't gone quite so into it, so here's a discussion I had on Palmtalk about it: https://www.palmtalk.org/forum/topic...en-cold-spots/
I guess this means that urban heat islands might bring my area of expertise closer to what they would be without the artificial cold instead of making them warmer than they would otherwise be.
And as for how this relates to a longer base than 30 years, that's a problem for the Southeast US. The official stations have started out in good downtown locations but end up getting moved to those very cold airports, meaning the normals can only go back so far before you end up mixing up 2 different places.
Perhaps longer normals will become more feasible many years into the future. But for now it's something I won't necessarily do.
So when climatologists use temperatures, do they use the daily highs, lows, or average them out?
Even if the night time lows are representative of natural vegetation rather than the heat island effect, what about the daytime highs? If the highs are significantly higher and if sensors have been moved to those locations or heat islands encroached on sensor station locations, wouldn't that still artificially yield higher recorded temps for those stations -- artificially above and beyond rural areas away from those heat islands?
So when climatologists use temperatures, do they use the daily highs, lows, or average them out?
That would be what they do. It is a good question.
Quote:
Originally Posted by Igor Blevin
Even if the night time lows are representative of natural vegetation rather than the heat island effect, what about the daytime highs? If the highs are significantly higher and if sensors have been moved to those locations or heat islands encroached on sensor station locations, wouldn't that still artificially yield higher recorded temps for those stations -- artificially above and beyond rural areas away from those heat islands?
I'll grant you that there is no study on this topic and so it is difficult to estimate. However, I do think the stations I know of have daytime highs that wouldn't be inappropriate in the natural vegetation from what I have seen (that is, colder highs somewhat due to the vegetation's same sun cover that keeps lows high, unless the warm air is advectional).
For example, here is this station in Valdosta: https://www.wunderground.com/dashboard/pws/KGAVALDO5
It has warmer lows than the airport on clear and calm nights, but does not get as hot on some nonadvectional days (for example, February 2023 monthly max was 29.5C vs 31C at the airport, I witnessed the month firsthand and there was a lot of sun heating vs advection).
In my opinion this is likely to be because of Valdosta's minimal urbanization, as well as the higher altitude of downtown vs the airport playing a much bigger role than heat island (high altitude increases Southeast US lows if it's not crazy high thanks to temperature inversions, but does nothing to increase them).
To be fair, though, I should have specified the uncertainty of the high temperatures.
Overall, I would say as long as the highs are a bit colder to go with the warmer lows, it's likely to be representative. But take this with a potential grain of salt until we have a study confirming it like for the lows.
I just got the email from Ed Mansouri which would fix all that. He said: ...
Gee, the president and CEO of the company of that sells the things thinks they're perfect. What a surprise!
Here's what the president and CEO of OceanGate said: "I've broken some rules to make this. I think I've broken them with logic and good engineering behind me."
How did that work out?
There are standards for how temperatures are to be measured. The same standards apply to the US Southeast as they do to everywhere else; everybody plays by the same rules. Just because this process makes your "list" seem somehow inferior to you doesn't mean you get to change the rules.
Quote:
Originally Posted by Can't think of username
@Jbgusa, sorry for derailing this thread - the heat island topic got way more out of proportion than I intended it to be. ...
You should be sorry because this was supposed to be about agreeing upon standards to measure the climate. Yet you derail it with the suggestion that you can create your own standards for your "list" based on the delusional belief that temperatures measured on the roof of a 1.4-acre 7-storey parking garage is "closest to what the natural climate of the area would be like". Nonsense.
Everybody plays by the same rules. Even FSU–no matter how many times they lose to Clemson.
Gee, the president and CEO of the company of that sells the things thinks they're perfect. What a surprise!
He most definitely does not think they are perfect. This is his full email - you'll notice he admits some italicized things (that are not temperature, but still) could be better.
Again, given the fact that the NATIONAL WEATHER SERVICE uses these and recommended them to me I'll be taking his word for it. Note also what he says about feedback in the second last statement.
"We concede in some cases we are collecting “good enough” data that ranges anywhere from marginally useful to very useful and we also provide some photographic/video data as well via our network of cameras.
The weather station you sent us a picture of? It’s on top of a parking garage so its wind readings are going to be elevated over ground readings.
As I recall its location was primarily chosen because of the view of the FSU Football Stadium its associated camera would have and I don’t recall we were accommodated with an alternative location at the time. So it’s just sort of stayed there over the years.
I’m not too worried about the temperature readings - I think it’s high enough off the structure where it’s not going to have its readings overly amplified by ground-based irradiance.
This project has often forced us to choose practicality over ideality yet we’ve continued to get feedback about how useful it is to folks in both the public and private sectors.
Hopefully we can move further in the direction of increasingly accurate standards-based data."
Quote:
Here's what the president and CEO of OceanGate said: "I've broken some rules to make this. I think I've broken them with logic and good engineering behind me."
How did that work out?
I wouldn't know. I'm not familiar with what you're referring to.
But presumably they did not:
-admit to the needed improvements of what they were doing
-get positive feedback
-have a government service that knows what they are doing use and recommend their stuff for what it's good at
like Ed Mansouri did.
Quote:
There are standards for how temperatures are to be measured.
That doesn't mean those standards are good standards. Standards that make the readings massively different from the natural environment - and it's been proven that they do - aren't good ones for very self-evident reasons.
Quote:
The same standards apply to the US Southeast as they do to everywhere else; everybody plays by the same rules.
But again, it doesn't mean they are good rules. The fact that the standards/rules lead to a notably different climate from that of the natural environment speaks for itself on whether they are or are not good.
If you wanted to have objectively good standards the weather stations would have to be put somewhere fully reflective of the natural environment. IE: a clear treeless space for the weather station is desirable if the natural environment is prairie, but not if it's thick forest.
I will give you that I find the analogy of FSU and Clemson entertaining.
Quote:
Just because this process makes your "list" seem somehow inferior to you
It's not even about that. It's about accuracy - and if you don't believe that, multiple of my favorite, well performing list places are artificially cold.
There are multiple people who would benefit from this MUCH more than me, for reasons unrelated to for-fun weather monitoring - ie: those who don't live at cold sink airports getting increasingly accurate weather forecasts. From "Modeling the Tallahassee, Florida Minimum Temperature Anomaly":
Quote:
Over the past fifteen years researchers have sought to understand the pattern of minimum temperatures in and around Tallahassee, Florida. During the cold season morning minimum temperatures can vary as much as 8°C between local sites. This variation creates a major concern for operational forecasters when minimum temperatures are expected to approach 0°C. The need to warn the public about damaging freezes is very important since a freeze will impact local farmers and other interests. Our
research has statistically modeled the Tallahassee minimum temperature anomaly to aide in predicting freezes and generally to improve the accuracy of operational minimum temperature forecasts.
And think about it. How much use are temperature readings several degrees colder than both the natural environment nor where people live anyway?
Quote:
doesn't mean you get to change the rules.
No, but I do think I should be scrutinizing them and proposing improvements if - like is the case - they are not exactly good rules.
Quote:
You should be sorry because this was supposed to be about agreeing upon standards to measure the climate.
At first I was and I said so, Jbgusa said I didn't have to be. Seems like a pretty done case - can we move on now?
Quote:
Yet you derail it with the suggestion that you can create your own standards for your "list"
I never said anything about making my own standards. I said that, with regards to the mention of heat islands making climates warmer than they are, that urban areas in my area of expertise are likely to be closest to the natural environment because they are not particularly urban and the airports are so disproportionately cold.
Quote:
based on the delusional belief that temperatures measured on the roof of a 1.4-acre 7-storey parking garage is "closest to what the natural climate of the area would be like".
Nonsense.
Tell that to the 3 studies and comparisons with areas around the world backing just that up. You know what the studies are so I'll just reiterate the comparisons part.
Joe from New Zealand says:
-Uncovered areas a few meters from forests can see dozens of frosts when covered areas see none at all
-The difference can be between 2-6C
Now consider that the Southeast US gets a quite large majority of radiational cooling frosts, and that it would be more disproportionate than New Zealand because the sandy soils get much drier than they would be without cover, further enhancing this. It's coherent with for example FSU, the radiational cooling difference can be 4-9C (and that is broadly consistent with the up to 8C from the ACTUAL STUDY).
And then there's the fact that the natural environment the station is at, has tall hills that drain cold air. An excellent comparison given the similar altitude discrepancy is a place like Augusta Daniel Field vs the rest of the metro - and Tallahassee would have even greater and more frequent discrepancies because being 3 degrees further south it is notably less affected by the advective cold fronts that reduce the cold air drainage. Even in a forest cleared spot, the readings would have high low temperatures for this reason - never mind the clay soil as well that radiates less than sandy soil.
We are probably going to have to either agree to disagree or leave well enough alone if we want this thread to stay on topic. That's what I will do.
using an inappropriately long period is going to tell you more about what the climate was rather than what it is. Eg if you're a farmer and use normals to decide the most appropriate time to plant your crop, 1991-2020 is almost certainly going to be a better guide than say 1931-2020. In other words, next year is more likely to me more similar to 1991-2020 averages than 1931-2020.
The only question is, similar to the previous one, how do we know it's 1991-2020 that will give us what we need to know and not, for example, 1996-2020 or 2001-2020? I'm not saying it isn't, but is there any evidence it is, or like the previous one is it just likely to be for human convenience?
Last edited by Can't think of username; 06-29-2023 at 10:08 AM..
He most definitely does not think they are perfect. This is his full email - you'll notice he admits some italicized things (that are not temperature, but still) could be better.
Again, given the fact that the NATIONAL WEATHER SERVICE uses these and recommended them to me I'll be taking his word for it. Note also what he says about feedback in the second last statement.
"We concede in some cases we are collecting “good enough” data that ranges anywhere from marginally useful to very useful and we also provide some photographic/video data as well via our network of cameras.
The weather station you sent us a picture of? It’s on top of a parking garage so its wind readings are going to be elevated over ground readings.
As I recall its location was primarily chosen because of the view of the FSU Football Stadium its associated camera would have and I don’t recall we were accommodated with an alternative location at the time. So it’s just sort of stayed there over the years.
I’m not too worried about the temperature readings - I think it’s high enough off the structure where it’s not going to have its readings overly amplified by ground-based irradiance.
This project has often forced us to choose practicality over ideality yet we’ve continued to get feedback about how useful it is to folks in both the public and private sectors.
Hopefully we can move further in the direction of increasingly accurate standards-based data."
I wouldn't know. I'm not familiar with what you're referring to.
But presumably they did not:
-admit to the needed improvements of what they were doing
-get positive feedback
-have a government service that knows what they are doing use and recommend their stuff for what it's good at
like Ed Mansouri did.
That doesn't mean those standards are good standards. Standards that make the readings massively different from the natural environment - and it's been proven that they do - aren't good ones for very self-evident reasons.
But again, it doesn't mean they are good rules. The fact that the standards/rules lead to a notably different climate from that of the natural environment speaks for itself on whether they are or are not good.
If you wanted to have objectively good standards the weather stations would have to be put somewhere fully reflective of the natural environment. IE: a clear treeless space for the weather station is desirable if the natural environment is prairie, but not if it's thick forest.
I will give you that I find the analogy of FSU and Clemson entertaining.
It's not even about that. It's about accuracy - and if you don't believe that, multiple of my favorite, well performing list places are artificially cold.
There are multiple people who would benefit from this MUCH more than me, for reasons unrelated to for-fun weather monitoring - ie: those who don't live at cold sink airports getting increasingly accurate weather forecasts. From "Modeling the Tallahassee, Florida Minimum Temperature Anomaly":
And think about it. How much use are temperature readings several degrees colder than both the natural environment nor where people live anyway?
No, but I do think I should be scrutinizing them and proposing improvements if - like is the case - they are not exactly good rules.
At first I was and I said so, Jbgusa said I didn't have to be. Seems like a pretty done case - can we move on now?
I never said anything about making my own standards. I said that, with regards to the mention of heat islands making climates warmer than they are, that urban areas in my area of expertise are likely to be closest to the natural environment because they are not particularly urban and the airports are so disproportionately cold.
Tell that to the 3 studies and comparisons with areas around the world backing just that up. You know what the studies are so I'll just reiterate the comparisons part.
Joe from New Zealand says:
-Uncovered areas a few meters from forests can see dozens of frosts when covered areas see none at all
-The difference can be between 2-6C
Now consider that the Southeast US gets a quite large majority of radiational cooling frosts, and that it would be more disproportionate than New Zealand because the sandy soils get much drier than they would be without cover, further enhancing this. It's coherent with for example FSU, the radiational cooling difference can be 4-9C (and that is broadly consistent with the up to 8C from the ACTUAL STUDY).
And then there's the fact that the natural environment the station is at, has tall hills that drain cold air. An excellent comparison given the similar altitude discrepancy is a place like Augusta Daniel Field vs the rest of the metro - and Tallahassee would have even greater and more frequent discrepancies because being 3 degrees further south it is notably less affected by the advective cold fronts that reduce the cold air drainage. Even in a forest cleared spot, the readings would have high low temperatures for this reason - never mind the clay soil as well that radiates less than sandy soil.
We are probably going to have to either agree to disagree or leave well enough alone if we want this thread to stay on topic. That's what I will do.
The only question is, similar to the previous one, how do we know it's 1991-2020 that will give us what we need to know and not, for example, 1996-2020 or 2001-2020? I'm not saying it isn't, but is there any evidence it is, or like the previous one is it just likely to be for human convenience?
There isn't anything special about 30 year periods ***1-***0, in particular those in the series 1961-1990, 1991-2020 etc. The series is commonly used because they follow on from the first widely used normals, which from memory were 1901-1930. Normals aren't simple averages, they involve controlling for things like changes in weather stations etc so they aren't typically released on the fly, eg you wouldn't typically be able to get a 1993-2022 normal. You'd typically have to wait for say the 2001-2030 normals to be published.
And why 30 years? Again there isn't some magic to 30 years that 27 or 34 etc etc doesn't have, it's partly convention and partly seems to be a compromise between being too short and therefore too influenced by year to year variability and too long and therefore too influenced by long term trends. Eg if we used say a 1931-2020 normal, about half the data would be from before the current warming period so today's average conditions would have drifted considerably from those normals for many locations.
The first normals were produced in the mid 1930s when most countries only had about 30 years of data, hence 1901-1931 for the first widely used normals.
And interestingly: Shorter-period normals, such as the new 15-year normals, are required by some sectors, such as energy, construction, etc., for applications that use normals to predict conditions in the near future.
The first normals were produced in the mid 1930s when most countries only had about 30 years of data, hence 1901-1931 for the first widely used normals.
And interestingly: Shorter-period normals, such as the new 15-year normals, are required by some sectors, such as energy, construction, etc., for applications that use normals to predict conditions in the near future.
Those looking for shorter periods of time are asking for trouble. The central limit theorem suggests that a sample size of at least 30 is needed.
Those looking for shorter periods of time are asking for trouble. The central limit theorem suggests that a sample size of at least 30 is needed.
I'm no statistician, but clearly some believe they are appropriate for their uses: shorter-period normals, such as the new 15-year normals, are required by some sectors, such as energy, construction, etc., for applications that use normals to predict conditions in the near future.
That would suggest that the lesser statistical robustness of a shorter period is outweighed by the rate of change ie 30 year old data is now too old to predict conditions in the near future, which seems plausible.
I'm no statistician, but clearly some believe they are appropriate for their uses: shorter-period normals, such as the new 15-year normals, are required by some sectors, such as energy, construction, etc., for applications that use normals to predict conditions in the near future.
That would suggest that the lesser statistical robustness of a shorter period is outweighed by the rate of change ie 30 year old data is now too old to predict conditions in the near future, which seems plausible.
15 years is an extremely short time period and the climate can flip dramatically in that period.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
