the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Jun 2024
The weather of 1740, the coldest year in Central Europe in 600 years
Abstract. The winter 1739/40 is known as one of the coldest winters in Europe since early instrumental measurements began. Many contemporary sources discuss the cold waves and compare the winter to that of 1708/09. It is less well known that the year 1740 remained cold until August and again in October, and that negative temperature anomalies are also found over Eurasia and North America. The 1737/40 cold season over northern midlatitude land areas was perhaps the coldest in 300 years, and 1740 was the coldest year in Central Europe in 600 years. New monthly, global climate reconstructions allow addressing this momentous event in greater detail, while daily observations and weather reconstructions give insight into the synoptic situations. Over Europe, we find that the event was initiated by a strong Scandinavian blocking in early January, allowing the advection continental cold air. From February until June, high pressure dominated over Ireland, arguably associated with frequent East Atlantic blocking. This led to cold air advection from the cold northern North Atlantic. During the summer, cyclonic weather dominated over Central Europe, associated with cold and wet air from the Atlantic. The possible role of oceanic influences (El Niño) and external forcings (eruption of Mount Tarumae in 1739) are discussed. While a possible El Niño event might have contributed to the winter cold spells, the East Atlantic blocking is arguably unrelated to either El Niño or the volcanic eruption. In all, the cold year of 1740 marks one of the strongest, arguably unforced excursions in European temperature.
- Preprint
(8880 KB) - Metadata XML
-
Supplement
(1108 KB) - BibTeX
- EndNote
Status: open (until 31 Jul 2024)
-
RC1: 'Comment on cp-2024-40', Philip Jones, 10 Jun 2024
reply
Overview
This paper is generally well written and informative. I have a few comments below, suggesting that more could be made of the CET and the long Dutch instrumental record, and possibly the longer Dutch proxy record.
The most informative part of the paper is the determination of the very cold periods during the 1739/40 winter across Europe and the later cold spring and summer seasons. The diary information you have found links well with the few long instrumental records. Putting the diary information into monthly and seasonal context is good. Often diary information is given without this context.
You later discuss possible causes. The poor harvests might have been due to severe cold in winter, the freezing of rivers, canals and lakes, which took a long time to thaw, so planting was delayed across much of the continent.
The sea froze off the eastern UK coast in the winter of 1962/63, but there doesn’t seem any mention of this for 1739/40, even though they had similar cold temperatures. 1962/63 is a more recent extremely cold winter that would be a good comparison. I see you have used winters from the early 1940s though. As an aside, the River Thames didn’t freeze in 1962/63 as after London Bridge was replaced in the 1830s, the river became tidal upstream of London. From the 15th century up to bridge replacement in the 1830s the Thames wasn’t that tidal above London Bridge. There is more on this in Jones (2008).
Specific Comments
- Line 32, Jones and Briffa (2006) is much more about the British Isles (maybe more so than Manley (1957) but it does provide long temperature series that are in the Netherlands and Germany. You have referred to Dickson (1997), but a couple of points about 1740 are worth mentioning: (1) It is referred to as the Forgotten Famine (the Potato famine just over a 100 years later is more well known) – you mention this a page later. Similar numbers to the 1840s left Ireland, about half to the North America and the rest to Britain. (2) During the 1739/1740 winter, the River Shannon froze over, something it’s not done since. It is unsure from the Dickson book where the quite wide river did freeze over. Unfortunately there is no instrumental data from Ireland in 1740, which makes the Shannon freeze over more important.
- Line 46, with respect to the warm 1730s in CET, the temperatures in this decade have only recently been exceeded (since the 1990s). The warm 1730s allowed for a dramatic increase in population of Britain and Ireland (more children survived), which might have made the impacts of 1739/40 much greater. Your later discussion doesn’t seem to fully note the contrast of the cold 1739/40 compared to the very mild 1730s, especially the mild autumns in the decade.
- An interesting aside about 1740 in the CET record. Using the 1961-90 base period, no year has ever had all anomalies of the same sign until 2023. Last year they were all positive. 1740 came close to being all negative, but for September.
- With Uppsala in Table 1 it is worth reading the 2002 paper in some detail. All the data before about 1739 come from a thermometer in an unheated room. If you plot the daily data, there appears to be a cut-off below which really cold temperatures in Uppsala were not measured. This does include the warm 1730s though. 1740 is the coldest year for CET, De Bilt and Berlin, but it does not appear that abnormal in Uppsala. The Q is whether this is real and was central and northern Fennoscandia not as cold as the other three locations. Uppsala would be well inside the Scandinavian High. This could be contrasted with the use of 1940 in Figure S1, which implies that northern Scandinavia may have been less cold? The less cold nature of Uppsala is mentioned later around lines 212-214.
- On line 153, you earlier referred to the winter of 1708/09. Better to do this again and to the one you’re discussing as 1739/40 and later in the paragraph. Later in the paper you refer to winters by the January, which can be confusing.
- The Dutch temperature series in van Engelen et al (2001) can be used to compare 1739/40 with 1708/09 and much earlier cold winters. The series goes back to about 1250 for almost all years, and is instrumental from 1706. This series classifies 1739/40 as 8, but gives a 9 to 1788/89 and 1829/30. For these winters values come from the long De Bilt record. 1708/09 is also an 8, but 1683/84 a 9. So the Dutch series wouldn’t undoubtedly say 1739/40 was the coldest, for the 18th As you point out though exactly when the coldest periods occur can have important effects on harvests and phenology, and 1740 was cold until September.
- The Dutch instrumental series back to 1706 is described in Labrijn (1945). I got a scanned copy of this from KNMI, and it is still available, but I can’t get google to find it. It is worth getting, as it is more informative than Manley (1974). Important parts of it are summarised in English.
- The Thames had its greatest frost fair in 1739/40 in terms of length. I can’t recall, though, the reference I read many years ago about this.
- Good discussion of the circulation types in the following few pages. You note the odd values at some sites in some periods. It is quite difficult to check these, when the sites are not that close.
- The temperature anomaly maps in Figure 6 seem to suggest that Uppsala wasn’t as cold relatively as the other sites, but did it influence all that is further north?
- Line 336-338. You refer to 1740 as being very cold, but then mention 1829? The winter 1829/1830 was also very cold. Here you refer to winters by the December, better if you gave both years.
- Just after this there are attempts to compare 1740 with respect to other years and winters, but this could also be undertaken with CET. Here 1740 was the coldest year (since 1659) at 6.9 deg C. The warmest years were 2022 and 2023, which where both 11.1 deg C, so only a 4.2 deg C difference. For CET, this is almost the same as you have in line 342, but you’re talking about a west European average?
- Lines 359-368, there has recently been a paper in Nature by Esper et al, which gives summer temperature estimates back 2000 years (for parts of the NH). Maybe worth comparing with that but this is summer. The coldest summer in the last 2000 years in the series was in 536, so much earlier.
- I’ve always wondered what might have caused 1739-40. The Japanese volcano is not large enough and the ice cores don’t show a major dust/acid layer. The ENSO influence is not that strong. It could just be a natural circulation occurrence!
References
Jones, P.D., 2008: Historical Climatology – a state of the art review. Weather 63, 181-186.
Labrijn, A.: The climate of the Netherlands during the last two and a half centuries, Tech. Rep. KNMI No. 102, Royal Netherlands Meteorological Institute, mededelingen en Verhandelingen, 1945. 2518
Labrijn, A.: 1945, K.N.M.I., Mededelingen en Verhandelingen no 49, Staatsuitgeverij,’s-Gravenhage.
Citation: https://doi.org/10.5194/cp-2024-40-RC1 -
RC2: 'Comment on cp-2024-40', Anonymous Referee #2, 25 Jun 2024
reply
General Comments
The manuscript is an interesting work on the great winter of 1739/1740 in Europe. It is well structured and raised. Text, figures and references are appropriate to the research objectives. The main conclusion is that this event was the result of natural variability, more than the consequence of radiative forcing (solar, volcanic, etc). The question here is what is the probability of this ‘special sequence of events’ (all of them related to atmospheric dynamics), that is, why this winter was singular, and it is very difficult to find other similar examples.
Specific comments
Table 1. I guess that ‘p’ is the measured pressure in each location (regardless its location above sea level), and ‘mslp’ is the pressure reduced to sea level. Correct?
Authors calculate indices for the NAO and EA patterns, but they affirm that ‘during the cold spell in January a strong high pressure system established over Scandinavia and at the same time a rather strong low pressure system developed over the northern Mediterranean’ (page 9, lines 245-246). This situation seems related to the positive phase of the SCAN pattern. In my opinion results would be more consistent if authors apply the same methodology used with NAO and EA to estimate the behaviour of this pattern during the studied period.
Figure 2 bottom. I don’t find in the text comments on this figure.
Figures 3 and 4. October? According to the authors ‘The fifth period noted in Fig. 2 is the month of October, which was persistently cold at most stations and which will be analysed in the following based on monthly charts’ (Page 9, lines 265-267). Why have you excluded October from the analysis in Figs. 3 and 4?
Role of ocean and land surface (pp. 18-19). Have you considered to study the possible role of the Atlantic Multi-decadal Oscillation (AMO)? The AMO is correlated to air temperatures and rainfall over much of the Northern Hemisphere, in particular in the summer climate in North America and Europe (Ghosh et al 2016; Zampieri et al., 2017). the AMO can also modulate spring snowfall over the Alps (Zampieri et al., 2013).
Technical corrections
Abstract, page 1, line 15. ‘The 1737/40 cold season’ Erratum? Won’t it be 1739/40?
References
Ghosh et a., l2016. Impact of observed North Atlantic multidecadal variations to European summer climate: a linear baroclinic response to surface heating". Climate Dynamics. 48 (11–12): 3547. doi:10.1007/s00382-016-3283-4.
Zampieri et al., 2017. Atlantic multi-decadal oscillation influence on weather regimes over Europe and the Mediterranean in spring and summer". Global and Planetary Change. 151: 92-100. doi:10.1016/j.gloplacha.2016.08.014.
Zampieri et al., 2013. "Atlantic influence on spring snowfall over the Alps in the past 150 years". Environmental Research Letters. 8 (3): 034026. doi:10.1088/1748-9326/8/3/034026.
Citation: https://doi.org/10.5194/cp-2024-40-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 668 | 143 | 21 | 832 | 32 | 5 | 5 |
- HTML: 668
- PDF: 143
- XML: 21
- Total: 832
- Supplement: 32
- BibTeX: 5
- EndNote: 5
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1