Weather Chat Thread

I've heard a lot of people say asphalt radiates tons of heat at night and thus stations placed near lots of asphalt will be milder at night than ones placed around grass.

I'm not an expert but I'm 99% sure that asphalt radiating lots of heat literally just means it cools down much quicker (if you disregard specific heat capacity). Weather stations have radiation shields so they shouldn't be affected by any heat radiation from asphalt. What they will be affected by though is the asphalt cooling down and subsequently cooling down all the air around it.

I saw a shoutbox conversation about places with April colder than December, or May colder than November. This is for the northern hemisphere, and above a tropical latitude, as it was about places with a lot of lag.

MORRO BAY FIRE DEPT
SAN NICOLAS ISLAND

Here's 24h mean, SMHI mean and max/min mean compared for Umeå airport 1991-01-01 to 2020-11-31 (should be quality controlled but there may be some missing data)



As you can clearly see, the max/min method yields a 0.5C difference compared to the 24h mean in July. The SMHI had a 0.2'C difference, with a maximum difference of 0.3'C in November. The average difference between SMHI and 24h mean was 0.05'C for the whole year, and 0.3'C using the max/min method.

Some stations have very even diurnal curves which means the max/min average will be close to the true average, but some stations have systematically skewed diurnal curves which means the max/min average will be different from the 24h mean average. 

In Umeå airport's case, the max/min average is colder than the true average in every single month on average. The SMHI method is slightly warmer in some months, and slightly cooler in others, but overall it evens out annually.

This is what each hour on the 23rd of July averages according to Weatherspark at Umeå airport and Trondheim airport. Weatherspark uses inconsistent time periods and data, so don't pay attention to the actual temperatures displayed as these are not accurate. However, the actual shape of the temperature graphs will be accurate as these don't really change much with different time periods.

Umeå airport:


Trondheim airport:



As you can see, the daytime curve and nighttime curve look much more even and similar to eachother at Trondheim airport than at Umeå airport where there's a very noticable plateau during the day, with a pointy curvature at night.

And this is why the max/min method isn't perfect for approximating the true average temperature.

For Umeå it was up to 0.5'C in difference, with a 0 3'C difference on average

Which makes it even worse. If every place had the same systematic deviation from the true average when using the max/min mean, then there wouldn't be an issue. The issue is that it gives differerent discrepancies from station to station.

Ultimately, max/min mean is like infoclimat. It's generally relatively accurate, and if you don't have access to true data (official data / 24h mean) it can be a quick and easy way to get a ballpark figure. However, when using it you have to take all the data with a grain of salt, especially when comparing different places, and you can't draw definitive conclusions on how much warmer or cooler some place is compared to another.

I'm not quite sure to what you're referencing with your comment. What do you not do if you use the same method for every location?

All the comments I've made have literally been about why they're not comparable the way you're describing it.

This is highs and lows at Umeå and Trondheim airport in summer according to Weatherspark (the actual values aren't completely accurate, but this is only an example so the only important part is that all the graphs are using the same data). As you can see, on August 8th, Trondheim obviously had a higher max/min mean (it might not be true irl, but again this is just an example).


But when we look at the average temp over time on the 8th of August, it's now clear that Umeå airport is actually warmer on average.


Not only is Umeå airport warmer on average for a longer period of time (Umeå airport is warmer for 13h each day, Værnes is warmer for 11h each day), the amplitude of the temperature difference is greater on average during the period where Umeå Airport is warmer. The total warmth can be seen as the integral of this temperature graph, or in layman's terms the total area underneath the graph. The difference in temperature will be the area between the two graphs. It's clear that the area between the red and blue line is greater between 5.30 AM and 6.30 PM, than the same area between 6.30 PM and 5.30 AM. This means the integral of the red line minus the integral of the blue line is positive and that Umeå airport has a warmer temperature on average. 

This is why the max/min method for approximating the average temperature isn't always comparable between stations, because in this example the max/min method yielded a warmer mean on August 8th at Værnes, when in actuality Umeå airport clearly had a warmer average temperature.

Again: the max/min method is an okay approximation of the average temperature, but it isn't definitive.

This isn't something that's up for debate. It's literally just pure fact by all rules of logic. I don't know if you're continuing to argue because you don't get it or because you think I'm paying too much attention to what you think is a very minor issue. If it's the latter then it's really unnecessary because on time and time again I've clearly stated that the max/min-method is fine but unequivocably imperfect. That's all I've said. 

If you genuinely just don't understand the fundamental flaws with the method, even after my myriad attempts at explaining it, then I don't know what to say. Either way I don't have anything else to add that I haven't already said and there's no point in my reiterating further.