March 2021 C.E.T. and EWP forecast contests

[BornFromTheVoid]

Latest update

 

The chances of finishing:
Close to average (6.2 to 7.2C) is to 31.1% (2 days ago was 29.4%)
Above average (>7.2C) remains at 2.8% (2 days ago was 2.8%)
Below average (<6.2C) is to 66.1% (2 days ago was 67.7%)

GFS for the 10th to 15th looks like averaging about 6.6C, or 0.3C below the 91-20 average

[Summer Sun]

4.2c to the 9th

0.7c below the 61 to 90 average
1.4c below the 81 to 10 average

___________________________

Current high this month 4.2c on the 2nd, 4th & 9th
Current low this month 3.8c on the 7th

[Timmytour]

A relatively chilly first 10 days of March. more commensurate with the 30 post-war years than the last thirty years....

 

[damianslaw]

8 hours ago, Timmytour said:

A relatively chilly first 10 days of March. more commensurate with the 30 post-war years than the last thirty years....

 

Based on the chart a much colder start to March than most years - rolling mean, nothing below the 3 degree mark.

[Roger J Smith]

Since 1981, these are values below 4.0 for running CET values at any point in March. 

1983 _ 3.8 on 1st 

1985 _ (4.0 on 22nd and 23rd, no running CET below that)

1986 _ 1st to 17th, min value -0.2 1st, -0.1 to 3rd, 3.4 by 10th, reached 3.97 17th 

1987 _ 6th to 28th, the min value was 2.3 13th to 15th, month ended 4.1

1988 _ 1st to 8th, min value 1.8 on 2nd.

1990 _ 1st and 2nd, 3.3 on 2nd. 

1991 _ 2nd only (3.4).

1993 _ 1st to 7th, min value 1.6 on 1st.

1995 _ 4th to 10th, min value 3.2 on 4th.

2001 _ 1st to 9th, min value 0.1 on 3rd.

2002 _ 2nd only (3.8)

2004 _ 1st to 5th, min value 0.8 1-2, also 3.9 on 12th

2005 _ 1st to 15th, min value 2.0 on 3rd.

2006 _ 1st to 26th, min value 0.1 4th. Still only 3.0 by 18th.

2010 _ 1st to 16th, min value 2.5 8th. 

2011 _ 1st to 9th, min value 3.1 on 3rd.

2013 _ 12th to 31st, min value 2.7 end of month. 

2016 _ 7th to 9th, min value 3.8 on 8th.

2018 _ 1st to 11th, min value -3.8 on 1st. Was 2.1 by 7th. 

__ __ a second cold spell produced a min in the running CET of 4.2 19th to 21st.

________________________________

summary, 18 of 40 possible cases had some running CET values below 4.0, a 19th hit 4.0 as a minimum.

The average duration was 10.0 days. Only four cases included any days after 16th. 

Just for clarity, note that these minimum CET values are running averages not specific daily values which would

likely have been lower, except for cases that were on the 1st, those would be daily CET means for the 1st. 

 

 

[damianslaw]

Thanks for the data above, March 2006 was very cold, first halves to 2005 and 2010 also. This March running close to 2011 and 2001, mmm 10 year gaps.. 

Emphasises oddity of March 2013, the first third was near normal, second half exceptionally cold. 

Marches 86 and 87 also very cold.

[The PIT]

Another big jump today Sunny Sheffield up to 4C -1.4C below average, Rainfall 28.2mm 46.5% of average

[Summer Sun]

4.6c to the 10th

0.4c below the 61 to 90 average
1.1c below the 81 to 10 average

___________________________

Current high this month 4.6c on the 10th
Current low this month 3.8c on the 7th

[Timmytour]

Comparing the three lots of 10 day periods in March upto 2020 (ignoring the 31st) .....the rolling mean CETs from 1907 are....

I wonder what it was in the atmosphere in the forty years between 1940 and 1980 that created the degree of differences that existed then that don't exist to anywhere near the same extent before or afterwards?

From the late 80s all periods show the relentless march upwards though it's notable that the last ten days is only just catching up with previous highs, the middle ten days is soaring to new heights and is remarkably very closely aligned to the last ten days, and though it too reached new heights,  the first ten days seems to have peaked just after the turn of the century.  But nevertheless it's noticeable that the first ten days of March is in a similar position to where the last ten days of March was 100 years ago 

[Roger J Smith]

I had a look at the data and noticed that rapid warming tendency from about 1936 to 1971 to put a more precise time interval to it.

By the way, are your data 11-year running means or some other interval? They are evidently not raw data (actuals). It makes no difference to the question being pondered, just wondering. 

As to cause and effect, the first thing that came to mind was more severe winters in the period, which would last to early March then leave the continental signal in place for rapid warming. However, that didn't entirely explain things as some of the more prominent contributions to the signal came with cold starts to March after relatively bland winters (for example, 1958, 1970). 

I wonder if coal dust in the atmosphere might have played a role, with the rapidly increasing sun angle in March perhaps able to cut through the chilling effects of that by about the 10th on average? Coal heating was largely phased out around the early to mid 1970s just about when this signal went away. Perhaps that same factor was mitigated before the 1930s by cloudier weather patterns which would have rendered it less potent. 

I am looking at my Toronto and NYC data sets to see if anything similar shows up in March trends there. Will post anything relevant.

[Weather-history]

March 2018 is a bit of con, although it is down as 4.9C, it was down to 5 days (sub 1.0C daily CET means) that made it that low. The first 3 days of the month and the St Patrick's weekend, take out those days and the rest of the month was about 6.1C

 

[Roger J Smith]

So I had a look at the same data analysis for NYC using the same ten day filters and the period 1907 to 2020, my data points are 11-yr running means, so in the graph attached, data point 1 is for 1907-17 and data point 104 is for 2010-2020. 

This shows a similar expansion of the three thirds of March temperature range hitting about the same time but not lasting as long ... by the 1960s at NYC mid-March was running cool and not getting any warmer than the first ten days had been. That tendency returned around the 1990s. Another oddity there is recently the last third of March was no warmer than mid-March. 

The most "expansive" March values on a year to year basis (greatest differences first ten to last ten) were in the smoothed data points closest to 1943-53, basically it was a pair of very warm Marches (1945,46) that drove this signal, and similar if muted results in 1948 and 1952, combined with a couple of notable swings from very cold to average temperatures in 1943 and 1950. Whether that has any transatlantic carryover or not, I don't know, but by 1965 which was a hefty contributor to the UK signal, a different regime of colder relative to normal trends had set in for quite a few of the Marches around then, meaning, the months started out near average and got colder than average, so there was very little upward movement of mean temperatures. That meant that the UK interval of "expansive" signals was cut short for NYC around 1948-58 although it had set in at about the same time. 

GRAPH: 11-yr running means of NYC temps (max + min / 2) in F deg (NYC Central Park)

1907-17 to 2010-20 data points

Edited March 11, 2021 by Roger J Smith

[feb1991blizzard]

5 hours ago, Roger J Smith said:

So I had a look at the same data analysis for NYC using the same ten day filters and the period 1907 to 2020, my data points are 11-yr running means, so in the graph attached, data point 1 is for 1907-17 and data point 104 is for 2010-2020. 

This shows a similar expansion of the three thirds of March temperature range hitting about the same time but not lasting as long ... by the 1960s at NYC mid-March was running cool and not getting any warmer than the first ten days had been. That tendency returned around the 1990s. Another oddity there is recently the last third of March was no warmer than mid-March. 

The most "expansive" March values on a year to year basis (greatest differences first ten to last ten) were in the smoothed data points closest to 1943-53, basically it was a pair of very warm Marches (1945,46) that drove this signal, and similar if muted results in 1948 and 1952, combined with a couple of notable swings from very cold to average temperatures in 1943 and 1950. Whether that has any transatlantic carryover or not, I don't know, but by 1965 which was a hefty contributor to the UK signal, a different regime of colder relative to normal trends had set in for quite a few of the Marches around then, meaning, the months started out near average and got colder than average, so there was very little upward movement of mean temperatures. That meant that the UK interval of "expansive" signals was cut short for NYC around 1948-58 although it had set in at about the same time. 

GRAPH: 11-yr running means of NYC temps (max + min / 2) in F deg (NYC Central Park)

1907-17 to 2010-20 data points

The relationship and synchronising or otherwise with UK Marches, could vary depending on PDO and AMO phases perhaps?

[damianslaw]

Perhaps not quite as mild looking mid-month as was looking a few days ago.. CET value by mid month quite possibly in the low 5s and no signal for anything especially warm thereafter. March 2021 is not going to go down as a particularly mild March at least.

[Roger J Smith]

This is an interesting graph that will appear in the Toronto data file about to be uploaded (in the climate change forum) ... after doing this analysis for Toronto I also did it for the CET ... running means of dates of record maxima (mean daily) for the year. 

How to read this ... data points are 10d running means of the years with record maxima. Where the graph goes lower, the records occur earlier (in the range of 1772 to 2020 since that's the period of daily data). The day number is at the bottom of the graph. For dates near NYD the running means overlap New Years (e.g. Jan 1 is Dec 27 to Jan 5). The ten day intervals are set up so that they run from d-5 to d+4 in each case. 

March goes from relatively recent records dominating mid-month to quite ancient ones near end of month (1777 in particular). That's the lowest point on the annual graph around days 85-90. The most recent portion of the records would appear to be late July-early August and quite generally Sept-Oct-Nov.

If there had been no warming trend overall, the equilibrium point would be around 1896 or (1772+2020)/2 but the actual equilibrium is closer to 1930.

[The PIT]

The charge upwards continues should slow a little for today and tomorrow then resume.

Sunny Sheffield up to 4.3C -1.1C below normal. Rainfall 34.5mm 56.9% of the monthly average

[Summer Sun]

4.9c to the 11th

0.1c below the 61 to 90 average
0.9c below the 81 to 10 average

___________________________

Current high this month 4.9c on the 11th
Current low this month 3.8c on the 7th

[Relativistic]

6 hours ago, Roger J Smith said:

This is an interesting graph that will appear in the Toronto data file about to be uploaded (in the climate change forum) ... after doing this analysis for Toronto I also did it for the CET ... running means of dates of record maxima (mean daily) for the year. 

How to read this ... data points are 10d running means of the years with record maxima. Where the graph goes lower, the records occur earlier (in the range of 1772 to 2020 since that's the period of daily data). The day number is at the bottom of the graph. For dates near NYD the running means overlap New Years (e.g. Jan 1 is Dec 27 to Jan 5). The ten day intervals are set up so that they run from d-5 to d+4 in each case. 

March goes from relatively recent records dominating mid-month to quite ancient ones near end of month (1777 in particular). That's the lowest point on the annual graph around days 85-90. The most recent portion of the records would appear to be late July-early August and quite generally Sept-Oct-Nov.

If there had been no warming trend overall, the equilibrium point would be around 1896 or (1772+2020)/2 but the actual equilibrium is closer to 1930.

Relative warmth in the second half of the year is a strange phenomenon that manifests itself in the monthly records too. With the exception of April, the first half of the year is dominated by older records, while the records from the second half of the year have all occurred since 1990:

 

January: 1921

February: 1869

March: 1957

April: 2011

May: 1833

June: 1846

July: 2006

August: 1995

September: 2006

October: 2001

November: 1994

December: 2015

 

Not something I have an explanation for.

[BornFromTheVoid]

Latest projections...

 

The chances of finishing:
Close to average (6.2 to 7.2C) is to 34.3% (2 days ago was 31.1%)
Above average (>7.2C) is  to 2.4% (2 days ago was 2.8%)
Below average (<6.2C) is to 63.3% (2 days ago was 66.1%)

GFS for the 12th to 17th looks like averaging about 6.2C, or 0.9C below the 91-20 average

[Roger J Smith]

EWP update, 20 mm to 10th, probably still below 25 mm and looking to add perhaps 15 mm with high pressure in control after this weekend. A rather dry outcome is becoming increasingly possible with the guidance now available (to 28th). 

This same high pressure regime presents two possible outcomes for CET. The more likely according to most current guidance is a rather cool regime with only one spike of limited warmth around mid-week next week, otherwise a blend of relatively cold and near normal spells with weak frontal passages between intervals of higher pressure. The second possibility is that the high actually settles a bit closer to Europe and anchors a prolonged warm spell, with a very warm phase as yet undetected by guidance. This is based more on the persistence of monthly trends from past few months. 

The more likely of the two outcomes (modified cold) would suggest we won't be out of the 5 to 6 range at any point after this weekend until perhaps very late in the month, so 5.5 to 6.3 a reasonable estimate. The second less likely possibility opens the door to outcomes in the higher 6 to low 7 range. 

Will run the provisional EWP scoring for 45 mm and let you know what that would do to the annual EWP scoring race. 

[damianslaw]

Mmm models suggesting rather average temps in coming days, and nothing especially mild for the foreseeable. 

[Timmytour]

On 11/03/2021 at 15:49, Roger J Smith said:

I had a look at the data and noticed that rapid warming tendency from about 1936 to 1971 to put a more precise time interval to it.

By the way, are your data 11-year running means or some other interval? They are evidently not raw data (actuals). It makes no difference to the question being pondered, just wondering. 

 

just noticed I left that information out!  Sorry Roger I'm using thirty year rolling averages

[Don]

May just may sneak in a below average month, albeit probably subtly!

Edited March 13, 2021 by Don

[Daniel*]

13 minutes ago, Don said:

May just may sneak in a below average month, albeit probably subtly!

Looks comfortably done further south.

[Daniel*]

This is interesting I feel there is a distinct chance.