Worth noting: NYC (Central Park) 2020 ties 2012 warmest year on record

[Roger Smith]

I have been doing some extensive research on historical weather trends at NYC and Toronto which both have long periods of record, and any other data from New England that extends back further such as Caswell's Providence RI journal (1831 to 1860). 

In the course of that, I was alerted to the existence of a daily data base and the official monthly temperature website for NYC Central Park. 

At some point soon, I can share that data base as expanded to show daily records and other derived products. 

In the meantime, it was made official (on their website publication) that 2020 had the same average as 2012 (57.3 F) to tie for warmest year on record. Three other years were 0.1 cooler. As you are no doubt aware November 2020 did the heavy lifting by finishing first as well, December could have pushed the year over the top but the mid-month snowstorm probably cut into the mean just enough to prevent that. 

Will provide a link to my deluxe data file soon, just in final edit now. Thanks to Don Sutherland who helped me acquire the data and alerted me to the website in question. We recently found a typo in that and queried it (May 2005) resulting in an edit on their part, what was 59.8 now reads 58.9 which makes that month even colder in the bottom portions of the May rankings. 

I could use some advice with one issue. For Toronto's data, I assumed an urban heat island began around 1881 and increased by 0.1 C deg a decade to about 1980 and then stabilized at that figure as urban heat islands tend to grow only marginally for very large cities, and in this day and age some improvements to air quality and the midtown parkland setting there probably counter any further heat island growth.

What would informed readers estimate to be the magnitude of the NYC heat island effect? I went with the same progression in my first draft, except that I assumed also that NYC having started to grow large before Toronto, already had 0.3 C heat island in 1869 when the weather station began recording, so that by applying the same corrections as Toronto, I am implying a heat island of 1.4 and saying that my adjusted figures are for comparison to a small heat island situation (0.3) rather than no heat island as the Toronto data come out. Or in other words, these adjusted figures would, if compared to each other, be like a comparison of rural Toronto to slightly urban New York City at each point. 

On my current schedule of adjustments, I think 1949 and also 1953 come out a bit ahead of 2012 and 2020 (which stay equal) since we only subtract 0.7 C or about 1.2 F from its mean annual temperature for comparison. 1990,91,98 and 2016 all come in just behind these tied for fifth warmest in the UHI-adjusted ranks. (any years tied after 1981 remain tied after adjustments). This is an issue I would like to address, has the heat island perhaps continued to increase slightly, or even decrease slightly since 1981? 

So am I in the ballpark of a good adjustment by saying 0.1 C for 1881-90, 0.2 for 1891-1900, 0.3 for 1901-1910, etc, ... 0.7 for 1941-50, 0.8 for 1951-60, 0.9 for 1961-70 and 1.0 for 1971-80, then finally 1.1 for all years 1981 onward? Is there any case to be made for any decrease? I have also considered a more fine tuned approach where I reduce dry months more than wet months (cloud cover differences). That may come in a stage two and it would need to be done for both locations. 

I assumed that 19th century heat islands were weak because of the lack of vehicles and home heating, but not zero because of albedo changes mainly. 

I have seen research into growth of heat islands and know how they diminish in their growth inverse-exponentially. Also for Toronto I had the benefit of some reasonably close rural stations with long periods of records to calibrate my estimates. For NYC the entire area is so urbanized that I find it difficult to estimate what the mean temperature would be at a Central Park like setting with no city around it (something that hasn't existed since about 1750 I suppose).

Any thoughts on this welcome. The long term goal of this project is to provide a climate record similar to the CET (Central England Temperature) series in which they claim to have factored out the urban effect although the three locations they use are fairly rural anyway. Toronto at 1.1 C is fairly conservative, obviously it's more like 5-8 C deg on a clear, calm night especially in winter, but the adjustment needs to be scaled to all weather types and on a cloudy, windy day the urban heat island there is probably no bigger than 0.5 C, if that. 

This is where the study resides at present time, on the UK equivalent forum "Net-weather" 

https://www.netweather.tv/forum/topic/93113-toronto-180-a-north-american-data-base-of-180-years-now-includes-nyc-1869-2020/page/2/

You'll see tables comparing NYC and Toronto using the assumptions listed above, but these could always be changed in a later version.

[Roger Smith]

Re the NYC data base, once again thanks to Don Sutherland for providing the basic data (columns A to E are his original file except that I've added recent months from CF6 official data) and everything else is work I've done to support the research thread on Netweather. I plan to post all that on a stand-alone website just so it isn't at the mercy of another host, going forward.

Unfortunately it is too big to upload to American Weather Forum. It is however downloadable from Netweather if you follow the link in the previous post.

As stated above, this excel file will be posted every three months after each meteorological season comes to an end, so if you're interested in the file, you won't necessarily have to update it yourself. 

Note that site navigation on the excel file is provided in a block located in columns GD to GR and rows 1 to 21 (this location is repeated in the intro panel upper left corner).

If you spot any errors (most likely to be color code omissions as the data base feeds directly into the tables) let me know and they will be fixed for the next edition. 

 

NEW YORK CITY (NYC) DATA BASE 1869 to end of FEB 2021 with record values and averages in tables ...

... available at this time only from Net-weather forum due to size issues, proceed to research thread as noted in previous post, link to the file is in the last post made by me in the thread on page 3. (it may not remain the last post until next upload)

 

[Roger Smith]

The Toronto file from the research is now also available on the same Netweather link (see above). 

This one resembles the NYC file but has additional features such as the tracking of record highs and lows from "starter" extremes in the period 1840-70, to the final values (which were usually later than that). Some days have had as many as a dozen records set from the starter record to the present time. Others did one change and that was it, a few managed to hold on to their original starter records (Aug 24, 1854 at 100 F for example). 

This is the second longest climate data base ever created (I believe) after the CET/EWP which go back into the 17th and 18th centuries. That is not to say that the Toronto data are the second oldest known, I have access to several data bases that are older but not continuous from then to now. 

Many of the daily temperature extremes at Toronto are in the same record spells that set NYC records. Sometimes they are one or two days earlier. Both locations had their highest temperature of record on July 9, 1936. In the case of Toronto that was one of three consecutive days but the mean daily temperature was higher on the 9th. 

Once again, I'd like to post these data bases here but they are too large to upload to American Weather Forum. My plan is to create a website with the data bases and other info so that eventually I won't be totally dependent on Net-weather for hosting them. 

[Roger Smith]

In the net-weather thread I have recently edited in these new data and a few corrections:

(a) March 2021 at 45.8 F (7.67 C) was t13 warmest with 1995, and Toronto (4.0 C) was t9 with 2020. 

(b) March 2021 at 3.41" was t60th driest with 1879. The trace of snowfall is tied with seventeen other years ranked t-4th. Only 1878, 1903 and 2012 recorded zero (no traces) snowfalls in March.

(c) A new daily record of 82 F was set on March 26, 2021 (replaced 76 1922).

(d) New high mins were set or tied on March 11th (52 broke 48 1898) and 26th (52 tied 1986). 

(e) An error was found in the max daily temps, Oct 15 should read 1956 rather than 1930 (84 F was okay).

(f) The 65 F record for Dec 28 was later tied in 2008. 

(g) The 60 F record on Jan 24, 1967 was not color coded. 

Some of these additions or errors can be adjusted if you downloaded the excel file. The corrections and additions will appear in the next instalment due around June 2, 2021. I am working on a log of record and notable daily highs, which tracks the evolution of the final set through a set of "starter records" for 1869 to 1900, adjustments to 1930 (about half the starters are broken in that interval), then any other new records in chronological order to the eventual record high (if not already set), then any notable close calls and a complete log of warmest days in the recent past (1991-2021 incl). This is about three-quarters finished and is the reason why I found the errors noted above. Hopefully this work will establish that all the listed records are correct. I plan to do the same for record lows where the starter records are going to do a lot better in terms of surviving later challenges. 

If you downloaded the Toronto file, there was just one error found since then, a ranking error, the details are in the net-weather thread.