August 2023 C.E.T. and EWP forecast contests

[Summer Sun]

17.0c to the 23rd

1.2c above the 61 to 90 average
0.2c above the 81 to 10 average

[damianslaw]

Interested to understand why CET is still measured against 61-90 means. When I wonder will it not? I think it used to be measured against 41-70 mean until about 1990. By that reckoning in 2010 should have moved to 71-00 and in 2020, 81-10.

[Roger J Smith]

The UK is the only country going that far back for its normal stats, as far as I know, which includes only three countries I know for a fact use 1981-2010 (Canada, Ireland) or USA (1991-2020 already). Ireland will adjust to 1991-2020 normals very soon according to their climate website. Canada has probably got 1991-2020 ready to go as well. 

Currently if you look at the standard U.S. product for climate summaries (the CF6) the daily reporting of to-date temperature anomalies is relative to 1991-2020 running means, no comparing to end of month normals there (until month ends). If you look at a Canadian product, you see the current average for the month and the end of month averages from 1981-2010, without an actual calculation. That is similar to what you would see on the Irish website under monthly data, today's average for the month, a few other years and the 1981-2010 average as normal. They say that will soon change to the 1991-2020 average. 

The ten-year convention means nothing in purely scientific terms, which is why I always include the most recent 30 year average available in my stats. When you think about it, the only reason we attach any significance to the number ten is that humans have ten fingers (usually). Otherwise we would not have base ten math, or a concept of decades, centuries, millennia etc. If we were endowed with six fingers on each hand we would no doubt have a base 12 arithmetic. Base ten is not what every developing human society chose either. I think the ancient Sumerians started out with a system based on sixty. 

The reason we have twelve months is of course related to the lunar cycle. Offhand I am not sure why there would be a seven day week unless it was approximately one quarter of a month? Will look into it. Maybe it evolved out of societal norms of how often one should rest from work, or how often one should worship gods. You could suppose the week is based on the seven days of creation but I think it's equally plausible that the first writing down of that story came after the emergence of the seven day week as a societal construct. Even if literally true, one could then ask, why would God choose a seven day week and not six or eight? 

Edited August 24, 2023 by Roger J Smith

[Addicks Fan 1981]

3 minutes ago, Roger J Smith said:

The UK is the only country going that far back for its normal stats, as far as I know, which includes only three countries I know for a fact use 1981-2010 (Canada, Ireland) or USA (1991-2020 already). Ireland will adjust to 1991-2020 normals very soon according to their climate website. Canada has probably got 1991-2020 ready to go as well. 

Currently if you look at the standard U.S. product for climate summaries (the CF6) the daily reporting of to-date temperature anomalies is relative to 1991-2020 running means, no comparing to end of month normals there (until month ends). If you look at a Canadian product, you see the current average for the month and the end of month averages from 1981-2010, without an actual calculation. That is similar to what you would see on the Irish website under monthly data, today's average for the month, a few other years and the 1981-2010 average as normal. They say that will soon change to the 1991-2020 average. 

The ten-year convention means nothing in purely scientific terms, which is why I always include the most recent 30 year average available in my stats. When you think about it, the only reason we attach any significance to the number ten is that humans have ten fingers (usually). Otherwise we would not have base ten math, or a concept of decades, centuries, millennia etc. If we were endowed with six fingers on each hand we would no doubt have a base 12 arithmetic. Base ten is not what every developing human society chose either. I think the ancient Sumerians started out with a system based on sixty. 

The reason we have twelve months is of course related to the lunar cycle. Offhand I am not sure why there would be a seven day week unless it was approximately one quarter of a month? Will look into it. Maybe it evolved out of societal norms of how often one should rest from work, or how often one should worship gods. You could suppose the week is based on the seven days of creation but I think it's equally plausible that the first writing down of that story came after the emergence of the seven day week as a societal construct. Even if literally true, one could then ask, why would God choose a seven day week and not six or eight? 

How do you expect the CET to be now @Roger J Smithas it seems to be moving albeit in a more sluggish way?      

[damianslaw]

Thanks for info Roger. Yes we are an oddity. Good to compare to 81-10 going forward I feel, more relevant baseline. With that some below average months are still happening such as July. Getting a below average month based on 1961-90 mean just seems a tall bar nowadays. 

[Don]

23 minutes ago, damianslaw said:

Thanks for info Roger. Yes we are an oddity. Good to compare to 81-10 going forward I feel, more relevant baseline. With that some below average months are still happening such as July. Getting a below average month based on 1961-90 mean just seems a tall bar nowadays. 

It is but we still manage to achieve at least one month below the 61-90 average most years, the last being December 2022.  In 2021, January, April and May were below the 61-90 average.

[reef]

It makes sense to use the most recent 30 year average. I'm not sure why the 1961-90 average is still used so much. There's no real reason to use 1981-2010 now either.

[damianslaw]

2 hours ago, reef said:

It makes sense to use the most recent 30 year average. I'm not sure why the 1961-90 average is still used so much. There's no real reason to use 1981-2010 now either.

Based on 91-20 mean, we've had 2 months below average this year, April and July, the latter comfortably so.

 August might not be far off the 16.6 mean average either. 

[Roger J Smith]

12 hours ago, Addicks Fan 1981 said:

How do you expect the CET to be now @Roger J Smithas it seems to be moving albeit in a more sluggish way?      

16.8 or 16.9 I think. EWP may increase a bit over my earlier estimate with latest guidance showing downpours on 31st across north, may move the bar closer to 83-85 mm. 

[cheese]

10 hours ago, reef said:

It makes sense to use the most recent 30 year average. I'm not sure why the 1961-90 average is still used so much. There's no real reason to use 1981-2010 now either.

Wasn’t the 1961-1990 period unusually cold too? Even colder than the preceding 30 year average? Very strange to continue using it when it has zero relevance to the modern day climate. 

 

Edited August 25, 2023 by cheese

[The PIT]

1 hour ago, cheese said:

Wasn’t the 1961-1990 period unusually cold too? Even colder than the preceding 30 year average? Very strange to continue using it when it has zero relevance to the modern day climate. 

 

Well it does climate change very well. If we used the modern day average it would show up less. I remember the met office saving temperatures had gone up by 0.8C since the sixties. I looked at my averages and they had gone up by 0.8C.

Anyway I uses the 91 - 2020 average as it seems more relevant.

Sunny Sheffield up to 17.1C +0.5C above average. Rainfall 52.6mm 76.5% of the monthly average.

Possible high point of the month?

 

Edited August 25, 2023 by The PIT

[LetItSnow!]

I think a combination of 1901-2000 and 1991-2020 should be used myself as the older average can be used to adress more extreme temperature anomalies while the 1991-2020 can fit it in a more contemporary context. An good example is April 2023, about -0.3degC below average using 1991-2020 but almost a degree above the 1961-1990 average. At least 1901-2000 would smooth out any decadal irregularities, such as the cold winters and cool summers of the 1960s and 1980s.

[Summer Sun]

17.0c to the 24th

1.3c above the 61 to 90 average
0.3c above the 81 to 10 average

[damianslaw]

A cool airstream looks like it will maintain itself rest of the month. Cool nights forecast as well, the CET could drop lower than expected. Outside of the CET zone, but today our high is 17.3 degrees with a low of 9 degrees, mean around 13 degrees only.. 

[The PIT]

Sunny Sheffield down to 17C +0.4C above average. Rainfall 53.6mm 77.9% of the monthly average.

Looking at the local forecast gives us a final figure of 16.8C so only slow drop to come to the end of the month. I guess for us anywhere between 16.5C and 16.8C is the likely ending zone.

[Summer Sun]

16.9c to the 25th

1.2c above the 61 to 90 average
0.2c above the 81 to 10 average

[Addicks Fan 1981]

I now think 16.5 or 16.6 will be the finishing number of this month.   This would equate to the summer overall at either being 16.5 or 16.6 rounded off, so in the same ball park as 1997 for our summer overall this year.   

[Don]

41 minutes ago, Addicks Fan 1981 said:

I now think 16.5 or 16.6 will be the finishing number of this month.   This would equate to the summer overall at either being 16.5 or 16.6 rounded off, so in the same ball park as 1997 for our summer overall this year.   

I think it will finish at 16.8C! 

No seriously, you are probably correct.  Has been a very different summer compared to 1997, even if it's around the same CET overall.

Edited August 26, 2023 by Don

[Snowyowl9]

The 1971-2000 average seems like the only one I m going to go for..

Mild 70s mild 90s...

[Dancerwithwings]

1 hour ago, Don said:

I think it will finish at 16.8C! 

No seriously, you are probably correct.  Has been a very different summer compared to 1997, even if it's around the same CET overall.

I’ll meet you both in the middle at 16.7c….now why would I think that

[Frigid]

49 minutes ago, Dancerwithwings said:

I’ll meet you both in the middle at 16.7c….now why would I think that

No I disagree, 16.6 is the number!

[Catbrainz]

Looking like CET will be in the mid or high 16s when August finishes. Pretty close to my prediction of 17c although my rainfall prediction is likely a fair bit too low. 

[Roger J Smith]

CET appears to be around 16.92 in two decimals so a bit of resistance built in to falls, today for example can stay at 16.9 (16.85) with as low an outcome as 15.2, so figure that if it falls to 16.8 with a slightly lower value, it will be even more resistant the next day. Any day that isn't a full unit below the running mean is probably not going to move the average (a unit being defined as number of days times 0.1 below the mean equalling 0.1, so by 30th, a unit relative to 16.7 is 13.7. To lose 0.1 a day, the last five days after today assuming it does move to 16.84 would be something like 14.2, 13.9, 13.7, 13.6, 13.5. I don't see it being quite that cool and you can estimate how many units are actually going to apply, if each day provides a half-unit drop then probably the outcome is 16.6. 

Latest EWP is around 70 mm now and GFS projection to end of month is close to 10 mm, so 80 mm is looking good; scoring from 78.1 mm in earlier table will not be adjusted until final value but at 80 mm it won't change very much. 

My estimate on the CET is also a finish on 16.6 or 16.7 now. 

Earlier I posted material on September cases of annual high daily means and maxima. This inspired me to add tables to the CET thread in the historical weather forum, outlining each year's highest max and lowest min, and it turns out that there are two more years with September high maxima that were not associated with year's high daily mean. I will be editing that post to add the information (I said that I didn't think there would be such cases but there are two anyway). 

[Addicks Fan 1981]

4 minutes ago, Roger J Smith said:

CET appears to be around 16.92 in two decimals so a bit of resistance built in to falls, today for example can stay at 16.9 (16.85) with as low an outcome as 15.2, so figure that if it falls to 16.8 with a slightly lower value, it will be even more resistant the next day. Any day that isn't a full unit below the running mean is probably not going to move the average (a unit being defined as number of days times 0.1 below the mean equalling 0.1, so by 30th, a unit relative to 16.7 is 13.7. To lose 0.1 a day, the last five days after today assuming it does move to 16.84 would be something like 14.2, 13.9, 13.7, 13.6, 13.5. I don't see it being quite that cool and you can estimate how many units are actually going to apply, if each day provides a half-unit drop then probably the outcome is 16.6. 

Latest EWP is around 70 mm now and GFS projection to end of month is close to 10 mm, so 80 mm is looking good; scoring from 78.1 mm in earlier table will not be adjusted until final value but at 80 mm it won't change very much. 

My estimate on the CET is also a finish on 16.6 or 16.7 now. 

Earlier I posted material on September cases of annual high daily means and maxima. This inspired me to add tables to the CET thread in the historical weather forum, outlining each year's highest max and lowest min, and it turns out that there are two more years with September high maxima that were not associated with year's high daily mean. I will be editing that post to add the information (I said that I didn't think there would be such cases but there are two anyway). 

If it is 16.7 then the CET for this summer will be 16.6, so classed as a hot and wet summer at the same time.   

[Don]

7 hours ago, Dancerwithwings said:

I’ll meet you both in the middle at 16.7c….now why would I think that

 

6 hours ago, Frigid said:

No I disagree, 16.6 is the number!

I think you are both likely to be closer than me!