Are there any signs of a global warming in Uppsala temperatures in 1739-1999? When did the Maunder minimum end?

An analysis by Timo Niroma, Helsinki, Finland.

Linear trend in Uppsala 1739-1999

Temperatures are in tenths of Celsius.

The regression is 4.88 + 0.0017 * x degrees C, where x = 1 year.
This means a rise of 0.017 degrees per decade.
Or 0.17 degrees per century.

Linear trends cut into five periods.

The ending and beginning temperatures do not necessarily coincide, because they are hypothetical temperatures based on the premise that the periods would have exactly linear rise or fall. However the curves are more or less rounded, as can be seen later.

1739-1801: The warming after Maunder Minimum, the coldest part of the Little Ice Age

The regression is 4.92 + 0.0063 * x degrees C, where x = 1 year.
Begins from 4.9 degrees and ends with 5.3 degrees.
This means a rise of 0.063 degrees per decade.
Or 0.63 degrees per century.

1801-1816: Deep dive into the Dalton minimum

The regression is 5.00 - 0.034 * x degrees C, where x = 1 year.
Begins from 5.0 degrees and ends with 4.5 degrees.
This means a fall of 0.34 degrees per decade.
Or 3.4 degrees per century.

1816-1867: It's a little warmer but the trend is much milder than during the Dalton minimum

The regression is 5.30 - 0.011 * x degrees C, where x = 1 year.
Begins from 5.3 degrees and ends with 4.7 degrees.
This means a fall of 0.11 degrees per decade.
Or 1.1 degrees per century.

1867-1930: The period of the greatest warming after the Maunder minimum or the Little Ice Age, the GLOBAL WARMING PERIOD

The regression is 4.47 + 0.011 * x degrees C, where x = 1 year.
Begins from 4.5 degrees and ends with 5.2 degrees.
This means a rise of 0.11 degrees per decade.
Or 1.1 degrees per century.

1930-1999: The warmest period after the Maunder minimum, but the warming trend is only 1/4 of what it was during the previous 63 year period.

The regression is 5.36 + 0.0029 * x degrees C, where x = 1 year.
Begins from 5.4 degrees and ends with 5.6 degrees.
This means a rise of 0.029 degrees per decade.
Or 0.29 degrees per century.


period          mean temp C    rise or fall per century
1739-1801 63yrs 5.1 degrees    +0.63 degrees
1801-1816 16    4.8            -3.4         Dalton minimum
1816-1867 52    5.0            -1.1
1867-1930 64    4.8            +1.1         global warming
1930-1999 70    5.5            +0.29

The warmest decades of the period 1739-1999 were the same as in Helsinki in 1829-1999: the 1930's and the 1990's, with temperatures of 6.0 and 6.2, respectively, in Uppsala and 6.0 and 5.9 in Helsinki.

There is one big surprise lurking in the 1700's data. The Uppsala temperature average in 1773-1779 was anomalously high (at that time): 5.8 degrees. What makes this discovery important, is that this temperature was exceeded only 160 years later, in the 1930's and again in 1990's, but all the decades from 1780-1920 and 1940-1980 were colder. How does the CO2-theory explain this? The solar activity level (as measured by sunspot activity) was at its highest in the 1700's. Beryllium in Greenland ice shows also that the decade of 1770's was also the sunniest (least cloudiest) of all the decades after the Maunder minimum end in 1700.

Sunspots (Wolf numbers): linear trend


The regression is 38.4 + 0.10 * x Wolfs, where x = 1 year.
This means a rise of 1.0 Wolfs per decade.
Or 10 Wolfs per century.


The regression is 33.8 + 0.34 * x Wolfs, where x = 1 year.
This means a rise of 3.4 Wolfs per decade.
Or 34 Wolfs per century.


The regression is 61.4 + 0.24 * x Wolfs, where x = 1 year.
This means a rise of 2.4 Wolfs per decade.
Or 24 Wolfs per century.

Summary: Uppsala temperatures compared with sunspots

period     / rise of temp   rate of spot incr            
             per century    per century 
1739-1999    0.17 Celsius   10 Wolfs
1867-1999    0.87 C         34 Wolfs     
1930-1999    0.29 C         24 Wolfs     

period     / average temp   average sunspots              
1739-1801    5.1  Celsius   51 Wolfs     warming 
1801-1867    4.9  C         42 Wolfs     cold
1867-1930    4.8  C         42 Wolfs     greatest warming
1930-1999    5.5  C         70 Wolfs     warm at both ends

Helsinki and Uppsala temperatures in 1830-1999 compared

mean annual temperatures during the period
years    Uppsala Helsinki difference 
1830-1849   4.9   4.1     0.8
1850-1869   5.0   4.0     1.0
1870-1889   4.6   4.1     0.5
1890-1909   4.9   4.5     0.4
1910-1929   5.1   4.8     0.3
1930-1949   5.6   5.5     0.1
1950-1969   5.1   5.1     0.0
1970-1989   5.3   5.3     0.0
1990-1999   6.2   5.9     0.3
2000-2004   N/A   6.2     N/A

From Dalton minimum to the small minimum in the 1860's there is about 5 degrees in Uppsala and 4 degrees in Helsinki. This difference of 0.9 degrees is the greatest during the whole period 1830-1999. The difference is much smaller, about 0.4 degrees, during the great rise from the 1870's to the 1930's (in Uppsala from 4.6 to 5.6 and in Helsinki from 4.1 to 5.5 degrees) and again in the 1990's.

But in-between, from the 1930's to 1980's, the annual temperatures in Uppsala and Helsinki are practically the same year after year. This includes the deep dive into 5.1 degrees in the 1950's and 1960's. In Uppsala this means a return to the temperatures of the previous century, in Helsinki we are one degree above the previous century and one degree below the 1930's and 1990's.

The great question, at the moment unanswered, is, has this oscillation from one degree difference thru no difference at all again to a 0.3 degree difference a real phenomenon, or is it caused by some bias in the measuring of temperatures or change in the local environment. Both cities lie at the 60 degrees Northern latitude, Helsinki is a seaside town, Uppsala is inland, Helsinki has several several hundred thousand more people than Uppsala and so on, so the differences are not easily explained.

Annual Uppsala and Helsinki temperatures compared on two occasions

Temperatures are in tenths of Celsius.


There is almost one degree warmer in Uppsala than in Helsinki, average being 4.9 degrees C in Uppsala and 4.0 in Helsinki. Correlation is .82. Minimum year in Uppsala has a temperature of 2.5 and in Helsinki 1.6 degrees. The year is the same, the famous catastrophic famine in 1867. The warmest year has a temperature of 6.7 degrees in Uppsala and 5.7 degrees in Helsinki. The year is again the same in both places, namely 1863. Remark that there is only four years between these extreme years. The 1860's is the coldest decade since Dalton minimum (in 1810's). As a sidemark, the sunspot cycle that ended in 1867, lasted one Jovian year. The coldest decades after 1930 have been the 1960's and 1970's. The only sunspot cycle of a length of one Jovian year after 1913 occurred in 1964-1976.


Not much difference here, average both in Uppsala and in Helsinki being 5.3 degrees C. Correlation is .96. Minimum year in Uppsala has a temperature of 3.3 degrees and in Helsinki 2.9 degrees. Both years belong to the famous cold years in 1940-1942. The warmests year has a temperature of 7.2 in both places. After the record year of 1934, the 7 or nearly 7 degree years come in somewhat regular intervals in both places: 1938, 1949, 1961, 1975, 1989.

Distance weighted best fit in Uppsala 1739-1999

Temperatures are in tenths of Celsius.

Best fit cut into three periods.


Note the ups in about 1760-1800 and 1820-1830. in between is the Dalton minimum and after that global chilling ending in the 1860's.

1867-1930: the greatest warming

There are three small ups and three small downs during these 64 years (the 22 year or Hale oscillation). Still the trend is very straightforwardly and powerfully up, the only small lacunae being around 1920, just before the jump into Modern Warm Period.

1930-1999: The Modern Warm Period

The deep dive culminating in the 1960's between the warm 1930's and 1990's is very clearly seen. Remark how accurately the 1990's follow the Sun's cycle (even better seen if the QBO is omitted). Remark also the steep rise from 1970 to 1975 and the drop in 1976. In 1976 ended the only long (one Jovian year) sunspot period after 1913. All others in the period 1913-1996 had a length of 10.3 years or 0.8 years shorter than the average since 1749.

How accurate is AD 1715 as the end of the Maunder minimum?

How simultaneous was the recovery of the Sun's activity and the beginning of the rising temperatures at the end of the Maunder minimum?

After prof. Eddy had in his article about Maunder minimum in Scientific American in 1975 rised that period from the half-forgotten and much neglected item into one of most discussed and debated items both in solar study and climatology, the interest about this somewhat mystical period has not lost a bit of its interest. Because we know a little more about the ending decades than the beginning of the MM, I begin with two facts. Eddy said that the Maunder minimum ended in 1715 which has since then been almost always as cited the end year of MM. How accurate a date this is, will next be discussed. I begin this discussion from the coldest decade of the Maunder minimum, which most probably is also the coldest decade during the second millennium, namely the 1690's.

Because 1690's clearly begins to the era of Maunder minimum being an almost spotless decade and people starving to death, especially in 1695-1696 when for example a third of the Finnish people died of hunger, it will be treated in a separate page, one about Maunder minimum itself.

Here will be discussed the first signs of global warming after that decade. The 300 year long global warming since 1700 has not been exactly linear, but the trend is clear. There have been three temporary drawbacks of cold since then, all however warmer than the previous cold period. The first is the Dalton minimum, which centered around the 1810's, the second centered around 1860's and the third around 1960's and 1970's.

As concerns the Sun, it swiftly awakened after the Maunder minimum. The two first cycles peaked in 1705 with 58 Wolfs and the second 1717-1718 with 60-63 Wolfs, or about the level of the Dalton minimum. The minimum in 1711-1712 had still two spotless years, but the minimum in 1723 had already value that is comparable with today's minima, or 11 Wolfs. After 1712 there has been only one spotless year, 1810 in the beginning of the Dalton minimum. Also the maximum in 1727 was almost exactly as high as in 2000 reaching 122.

So some data we have about the sunspot activity from the two cycles in 1700-1723, which very much remind the Dalton minima cycles in 1798-1823 as their height and length. So with thight criteria - virtually no sunspots or annually either none or at most about ten sunspots, we should say that the Maunder minimum ended about or exactly 1700. If we accept a low intensity - the Dalton-like continuum after the Maunder minimum as part of it, the Dalton minimum ended only in 1723.

Sadly the first regular observations of temperature began only in 1722 in Uppsala, but are very far from certain until 1739. From the period 1700-1721, that just is critical, because of the awakening of the Sun, there is nowhere regular temperature data. To catch something, we could begin by watching how the temperature behaved during the Dalton minimum.

If this pre-Dalton behaved like the actual Dalton, the temperature would have begun its rise with the rise of the Sun's activity or about 1700 peaking 1703 and then being low again around the low maximum (typical for such maxima) or 1704-1707. The annual sunspot zero years of 1712 and 1810 would also mean cold (as always). After that, during the second cycle, the Dalton minimum temperature was taken into the QBO cyclicity which was cut in 1817-1818 with a deep dive in 1819-1820 and the ending rise in 1821-1822. Are there any data from this period in 1700-1721 which would confirm or cancel the hypothesis of Dalton-like behavior not only in the Sun but also in temperature?

G. Manley in Meteorological Magazine, Vol 90, 1961 says that there was a noticeable glacial advance in 1715-1720 in North. This would mean that the retreat may have begun soon after 1695 and taken back during the latter part of this pre-Dalton. "Snow frequency was about as twice in 1691-1706 that we should expect today" says the same source. If true, this gives an added phenomenon to add to the end of the Maunder minimum. If we could rely on the Uppsala temperatures, there is at least a hint to warmer times in 1722-1738.

Comments should be addressed to

More information about global warming

For more information about sunspots:
Goto or

For more information about global warming, especially in Helsinki perspective:
Click global warming

For more information upon this server:
Goto index
For information about the Maunder minimum in the 1600's:
Goto Maunder minimum