CC3 -- Why I believe that a third option is needed for climate change

Hello, we are back (for a while) ... I asked if I could post an update on my bogged down project to get Toronto data fully up for discussion over on Net-weather. It has been a marathon and while I knew it was going to take weeks, a few issues came up with data since 2003 when the precip was being measured in two different styles and time frames, and even worse since mid-2017 when the second option disappeared from the website leaving me with no daily snowfall numbers at the location (but daily snowdepth reports together with precip, and a fairly dense net of other climate stations in Toronto as a guide).

I am hoping that when I contact Environment Canada with the results (in about a week perhaps) they might be able to enlighten me about the details of the timing issues (suspecting some precip data changed from the long-standing calendar day to 0800h-0800h reporting) as well as a few other questions that arose about early practices (1840 to 1848 mostly) that sometimes appear as though accumulated rainfall gets into the data looking like daily specific values.

Anyway, the data posting is very near the finish line (I am basically in sight of the stadium anyway) with just daily precip records for September to December left to post, plus a summary already available of dry spells in the 180 years. Then all the data with monthly rankings will be up and available, and I can link to my massive excel file of the data which has many graphs to illustrate various trends and outcomes.

There will also need to be posts created for modified temperature rankings with the urban heat island removed from the trend. I have made a start on that in house and found that while it narrows the gap between recent and not so recent warmest months and in some cases brings back into top spots some of the eclipsed 1920-1950 era values, it only tweaks the order overall because most of the colder half of the rankings were from pre-heat island times and remain below the adjusted values of the more recent cold months, although, and somewhat remarkably, the coldest December would now be 1989 (third in raw data), the coldest Febs would be nearly a three-way tie of 1885, 1934 and 2015, and the coldest January top five would see 1994 moving from fourth to second coldest. This would be after a relatively modest urban heat island progressive downward adjustment starting at zero in 1881 and working its way up to 1.2 by 1980 (not increased thereafter as large city urban heat islands flatten out once the city expands beyond a 50 km diameter). To set the urban heat island's intermediate values I have compared two long-period outlying and entirely rural locations with the Toronto data to establish trends, and checked the research literature on magnitudes of heat islands (something that I studied in my university days and even made a contribution towards) -- the general underlying theme is that the heat islands take off with a fast start at around 10,000 to 50,000 populations then slow down in reverse exponential form after about 100,000 (which Toronto had reached before 1901) but there's the issue of how older data might respond in the earlier stages of modern urban heating parameters (different styles of home heating, fewer vehicles per capita on roads etc), so have come up with a somewhat delayed total response model that blends the traditional reverse exponential with a more linear paradigm to account for the increase in vehicle traffic and generally higher ambient home heating standards (warmer houses and businesses can emit more heat into the heat island).

Picture for example the typical Toronto neighbourhood around turn of century (1900), houses mainly heated by coal or even wood stoves, not oil or natural gas, and few if any people owning a car, just a few trolleys and cars on roads, so very few paved surfaces. On the other hand, Toronto has a considerable "urban forest" and I suspect that it is largely second growth with perhaps an interval around 1880 to 1900 when it was substantially removed at least within a few miles of the weather station. There was no university campus until around 1890, so I suspect there may have been a lot of open space with construction sites not too far from the pre-1908 location which would be 1.2 km south of the later (1908 to 2003) location. This may be partly why there's a bit of a spike in annual maximum temperatures around 1916 to 1922 which is actually the warmest seven year average of that statistic (1933 to 1939 very similar, later spikes in 1950s and 2010s not quite as high). I will have that graph available for posting. The interval from 1978 to 1986 had almost as low an average for the statistic as the coldest stretches of the 19th century. The maximum never broke 35 C (95 F) from 1978 to 1986 except for one day at 96 F in 1983. The long-term average of annual maximum works out to 34 C (93 F) and some intervals exceed 36, but it retreated to about 32 in the period 1978 to 1986.

The adjusted ranks of monthly and annual temperatures can be better used than the existing raw data but even that tends to show quite clearly that the main episode of warming at Toronto took place around 1890 and rather quickly, almost a phase change during a short period of time perhaps 1889 to 1893 after which the data start to look very similar to the modern period.

Some other parameters show trends that are probably at variance with conventional climate change assumptions but may still be significant. In general terms, rainfall started out very high in the 1840s, drifted down to its lowest point by the 1870s (but that partly because this was the coldest interval so more winter season precip fell as snow than rain), bounced back to long-term average oscillating tendencies from about the 1880s to 1920s, went into drought mode in the 1930s, recovered from that back to the long-term average 1950s-1970s, hit a notable wet spell mid-1980s, somewhat drier through 1990s into mid-2000s decade, then has been steadily approaching a return to the wetter mode of the 1840s in most recent years. While it has some way to go yet to rival 1840-44 (the wettest five year interval) it is clearly wetter in recent years than what I recall from my younger days when I lived in that region.

Unfortunately the station stopped recording sunshine hours some time around the 1980s, still searching for any data to extend a printed document I have of sunshine hours 1882 to 1967. For some unknown reason, EC with their very extensive climate data base ignore the existence of sunshine hours as if nobody had ever measured them. There used to be tables of daily and monthly values in the printed forms of the monthly climate bulletins that are essentially the source of the internet data base otherwise. And I suspect partly from my own memory of weather trends that sunshine hours peaked around 1962 to 1964 and steadily declined after that back to the cloudier mode that was evident in the 1880s and 1890s. This may have some connection to jet contrails spreading out, I know this has been discussed in the literature as a possible cause of enhanced cloudiness.

Another trend that is fairly stark is that recent temperature increases are entirely nocturnal (at Toronto anyway). There has been a 2:1 shift in ratios of record high mins to record high maxes. The number of new record high mins since 1970 is almost two thirds of the data set, whereas the number of new record highs is barely one third and is also barely above the random expectation for 53 of 180 years of data. At the risk of generalizing too coarsely, the situation seems to be that after a burst of multiple record highs set in the high-variability 1970s, the pace slowed to a crawl in that colder spell, resumed for a time (1987-91 set a fairly robust number), and has oscillated around random since that. And I could say that almost entirely the new records are either (a) in the winter shoulder seasons Nov-Dec and late Feb-Mar, or (b) picking off the weakest of the herd of older records. I can only find three "older" record maxima that I would consider robust that have been edged out (in each case by about 1 degree). But with the record high minima, it seems that any mid-level warm spell nowadays will threaten the existing set of record mins. It has become typical for recent heat waves (which are a regular feature of the climate) to produce days that are 2-4 deg below record values and overnight lows that are close to the record values. In general terms, the old heat wave standard of about 35/21 (95/70) has turned into 33/23 (92/73) with more cloud and higher humidity levels. The humidex values are probably similar. There are exceptions but that seems to be the background theme of the modern summer climate. This blends well with the observation that there has been an increase in summer rainfall too, not by number of days though (that statistic seems to be constant).

Despite the above, the actual number of new record high rainfall values is very close to random expectation and the maximum values are almost identical to the previous ones. For example, the highest one-day rainfall in the data set was 97 mm in July 1897; now the second highest one is 96 mm from July 2013. The highest January rainfall of about 63 mm (1843) was recently duplicated in January 2020 (but before that had not been approached within 20 mm). The same story is told by rainfall data in other months. I am working my way through the autumn months now and the number of extra tropical storm events associated with record daily rainfalls is remarkable even going back well into the mid-19th century. This trend has continued but I certainly see no evidence that it has increased -- it seems that even in the colder climate of the 19th century, when other pressure pattern studies would show clear evidence of a more southward jet stream position than in the 20th century, the same number of tropical systems managed to find their way to the Great Lakes region, and would then dump record rainfalls (many of them are in the 50-100 mm range). The total number over the entire data would be upwards of 50, and I have less evidence available for the 1840s than other decades. So by September almost half the daily records have some association with tropical remnants. This continues to mid-October then dies off by early November. The number early in the season is small, two cases in June, none detected in July, then a sprinkling in August.

Winter snowfall is certainly in a declining mode. Toronto is not in the climate zone of the U.S. northeast coastal cities where it is sometimes claimed that recent winters have seen increases (and those are due to climate change). Neither of those claims really survives a rigorous test anyway, but what is certainly true at Toronto is that with the notable increase in winter shoulder season temperatures has come a reduction in snowfall then also. The reduction is less evident in later December and January or the first half of February. Overall the number of new snowfall record values is about two thirds of random expectation since 1970. Despite that, there was a conflicting signal around the period of 1968 to 1976 for increased numbers of unseasonable snowfalls. New marks were set for latest in season 25 cm and 10 cm snowfalls (Apr 1975 and 1976), and earliest 10 cm (Oct 1969). So the 1970s are not really claimed to be a mainstream portion of the recent warming anyway, but what apparently took place in the 20th century warming was a relapse to mid-19th century (with a full urban heat island modification factor) for about a decade and a half, then with the stronger El Nino events around 1982, 1990, 1998, 2006 and more recently, pulses of warmth that have brought things back to about where they stood in peak warming times like 1916-22, the 1930s and 1946-59.

I am starting to think of the 1960s now as a period when the mid-20th century warmth began to flicker a bit, when it was "on" there could be record values even up to monthly maxima (May, June and October all set records in the sunny years of 1962-64). But the warming was not on as frequently as in the early 1950s for example. Then it went into high-variability mode on an epic scale (some of the ups and downs of the 1970s are stupendous and I was lucky to have lived through them in more ways than one). I've thought more than once that if we just had the 1970s over again the climate change lobby would go berserk with angst, if they think this rather damped down recent climate is pushing the limits (which I consider a faulty assumption based on ignorance of the data, hence this project), then they would really "freak out" if the 1970s came around again. Consider for example, April 1976, a week of record warmth followed by a heavy snowstorm.

Nowadays if one day touches 20 C in April there is no end to the "we will all die" chatter, and if it were to snow even an inch I am sure people would be told that it was out-of-control climate change at work. From all available evidence, I would have to imagine that the average politician, media reporter, or even climate change student, would score about 10 out of 100 on any quiz of true or false on past climate statistics. I imagine a quiz with questions like this: True or false, the majority of days above 35 C were before you were born.

You would need to be 81 years old (in Toronto) to answer "false" to that question.

Not sure if we'll keep the thread open or not, but I did want to mention that while I value open discussion and free speech, at the same time I saw no useful purpose to much of the off-topic intrusions (the sidereal day stuff in particular) and if we do continue I think anything that far off the straight line should be eliminated. In the absence of a lot of material on topic from myself, the discussion rather predictably reverted to the default values of AGW vs skeptic rhetoric which might be worth a separate thread somewhere else; the whole point of my new perspective being that perhaps the best response to the question of climate change is to assume warming and work from there. I don't think there's anything meaningful that we can do to alter the outcome because I feel like two thirds or more of the climate change response is naturally driven anyway so that our attempts to modify our portion of the complex outcome will be negligible even though perhaps harming the economy by trying (as we are seeing in Canada recently, overly hysterical responses to the situation are crippling economic planning and ripping apart a rather precarious social fabric with no winners, just each faction losing). The experience of the former Yugoslavia can teach us that in a breakdown of civil society and widespread factional fighting, there can be an outcome where everyone loses and nobody wins.

If that's what the AGW brigade want, then so be it. It's not what I want.

It has been mild in a lot of places alright, Toronto has just finished its 10th mildest winter out of the 180 in my study. When the urban heat island is factored in (those tables still to come) I think it will drop back to about 15th as some of the ones it edged out were a long time ago and were not beaten by even half the magnitude of the heat island (for example 1889-90 will probably move back ahead of this winter in that more relevant comparison).

Where I live, the temperatures have been fairly close to average and snowfall about half to two-thirds of normal. We are up high enough in the mountains that any warming trend is hard to spot since a bit of additional warmth is often likely to generate more precipitation, but this past winter has been relatively dry as well. Nowhere near record-breaking here though, just a touch above average.

The eastern U.S. had a mild winter, the desert southwest was a bit below normal at times. This seemed to be a weaker and blander version of some past winters with a deep western trough and cut-off lows that bring unusual snowfalls to the southwest. There was one episode of that but it didn't really build up very much.

If my theory that the circulation is influenced to some extent by the shifting magnetic field, then a lot of the discussions about trends will be pointless because the changes underway could lead to climate patterns we have not seen in modern times. I think this might become more apparent in 20-40 years from now, if the NMP continues to drift towards northeast Siberia. It would be a rather warm climatic influence to have the polar vortex trying to form more often in that region. Eastern Asia might turn colder but most other regions would probably see a warming influence from that. I wonder if the NMP was in that position during the MWP. Newfoundland, Labrador and Greenland were all relatively warm in that era and it stands to reason that if the polar vortex distorts westward (I wouldn't expect it all to shift that way as I think this is a second-order influence) then southerly flow is encouraged in the Atlantic basin generally speaking.

I am hopeful that some people will take some time to look over the material in that Toronto study and offer some thoughts about what they see. I have grown up with that information (the study uncovered some new information for me, but most of that stuff is as familiar to me as the CET or past Irish weather might be to some of you). I had partly lost touch with trends since we moved west in 1995, and this helped me to refocus on a subject that used to be the core of my research in the first decade or so.

In this field, you can very easily suffer from data overload. I need to step back a bit from the enormous mass of data and do some creative analysis now. What can we learn from all this information? Maybe there are trends or tendencies that don't leap off the page. But I think it's hard to miss the fact that Toronto has always had quite a variable and dynamic climate that keeps going from extreme to extreme on a fairly regular basis. I don't see any real evidence that this is increasing. The high water mark for variability was probably the 1970s. The period around 1900 to 1920 was quite variable also. Winters seemed to be either/or with not many stuck in the middle.