The regime shift of the 1920s and 1930s in the North Atlantic
By Paul Homewood
The warming of the Arctic in the 1920s and 30s is well documented, despite attempts to wipe it from the temperature record. But it is always good to come across another paper, even though this one dates back to 2006.
During the 1920s and 1930s, there was a dramatic warming of the northern North Atlantic Ocean. Warmer-than-normal sea temperatures, reduced sea ice conditions and enhanced Atlantic inflow in northern regions continued through to the 1950s and 1960s, with the timing of the decline to colder temperatures varying with location. Ecosystem changes associated with the warm period included a general northward movement of fish. Boreal species of fish such as cod, haddock and herring expanded farther north while colder-water species such as capelin and polar cod retreated northward. The maximum recorded movement involved cod, which spread approximately 1200 km northward along West Greenland. Migration patterns of “warmer water” species also changed with earlier arrivals and later departures. New spawning sites were observed farther north for several species or stocks while for others the relative contribution from northern spawning sites increased. Some southern species of fish that were unknown in northern areas prior to the warming event became occasional, and in some cases, frequent visitors. Higher recruitment and growth led to increased biomass of important commercial species such as cod and herring in many regions of the northern North Atlantic. Benthos associated with Atlantic waters spread northward off Western Svalbard and eastward into the eastern Barents Sea. Based on increased phytoplankton and zooplankton production in several areas, it is argued that bottom-up processes were the primary cause of these changes. The warming in the 1920s and 1930s is considered to constitute the most significant regime shift experienced in the North Atlantic in the 20th century.
All of the factors were present then that we see now – warmer sea temperatures, reduced sea ice, enhanced Atlantic inflow, and ecosystem changes.
The Introduction offers more detail:
In the 1920s and 1930s, there was a dramatic warming of the air and ocean temperatures in the northern North Atlantic and the high Arctic, with the largest changes occurring north of 60oN (Rogers, 1985; Polyakov et al., 2003; Johannessen et al., 2004). This led to reduced ice cover in the Arctic and subarctic regions and higher sea temperatures. Jensen and Hansen (1931) and later Jensen (1939, 1949) documented the expansion of Atlantic cod (Gadus morhua) and halibut (Hippoglossus hippoglossus) along the west coast of Greenland in response to the changes in the ocean climate. Other species were also observed to have undergone significant abundance and distributional changes. This was a clear case of an environmentally driven ecosystem response that became of paramount interest for fishery researchers at the time. This interest lead to the first scientific meeting by ICES on climate change held in 1948 at Copenhagen (ICES, 1949) entitled Climate Changes in the Arctic in Relation to Plants and Animals. Ahlmann (1949), in his introductory address, noted that the warming had broad geographic extent with significant effects in the region: increasing air temperatures, receding glaciers, decreasing Arctic ice extent and thickness, decreasing water levels in lakes through increased evaporation, and high sea-level elevations due to melting ice.
This warming event was associated with atmospheric changes causing increased transfer of heat from low to high latitudes (Brooks, 1938; Ahlmann, 1949; Rogers, 1985). Indeed, increased southerly winds pumped warm air into the northern North Atlantic and also into the Arctic. Overland et al. (2004) showed that the Icelandic Low was located farther to the east than usual in the 1930s with the result that Northern Europe was subsequently warmed by winds from the southeast. This is in strong contrast to its normal warming from the southwest associated with a positive North Atlantic Oscillation (NAO) phase. In the Northwest Atlantic, a high-pressure system over Greenland caused warm southerly flow over Baffin Bay (Overland et al., 2004). Bengtsson et al. (2004) proposed that the temperature increase was related to enhanced wind-driven oceanic inflow into the Barents Sea with an associated sea-ice retreat. Through feedback mechanisms, this in turn generated and enhanced the cyclonic low pressure in the region and created a strong surface heat flux over the ice-free areas, a mechanism previously proposed by A ˚ dlandsvik and Loeng (1991). Modelling studies suggest that these changes might be caused by internal, non-linear dynamics of the atmosphere (Delworth and Knutson, 2000; Bengtsson et al., 2004).
Interestingly Drinkwater states (my bold):
Given the earlier reviews noted above, why is a new review needed? First, new information has become available, most noticeably on phytoplankton and zooplankton, as well as on the response of fish stocks such as capelin (Mallotus villotus) and herring (Clupea harengus). Second, our understanding and knowledge of some of the physical and biological processes have increased so we are in a better position to determine what likely happened. Finally, while many fisheries scientists working during that era were familiar with the event, many of today’s marine ecologists and fisheries scientists have either forgotten or do not know about it. This needs to be rectified given the important lessons it can teach us about what to expect under future climate change.
He describes the physical changes:
In the North Atlantic (Fig. 1), air temperatures through the latter part of the 19th century and the early 20th century were relatively cool compared to years since then. During the 1920s, and especially after 1925, average air temperatures began to rise rapidly and continued to do so through the 1930s (Fig. 2a–f). Mean annual air temperatures increased by approximately 0.5–1 C and the cumulative sums of anomalies varied from 1.5 to 6 C between 1920 and 1940 with the higher values occurring in West Greenland and Iceland. Farther south along the east coast of Canada and the northeastern United States, as well as in much of southern Europe, no such temperature increase occurred (Fig. 2g–h). However, even farther south, for example around Cape Hatteras and at Funchal in Madeira off the west coast of Africa, the large temperature increase was also observed (Fig. 2i–j). Throughout the remainder of this paper, only the changes associated with rise in temperatures in the most northern regions of North Atlantic are considered.
Through the 1940s and 1950s air temperatures in the northern most regions varied but generally remained relatively high (e.g. at Nuuk in West Greenland, Fig. 3). Thereafter, there was a rapid cooling trend with the exact timing of the decline varying spatially. In the Northwest Atlantic, warm conditions remained through most of the 1960s whereas in the Northeast Atlantic they began declining slightly earlier. The high temperatures recorded during the warm period from 1930–1960 match, and in some cases exceed, the present day warming (Johannessen et al., 2004).
And reminds us that they were intimately connected with sea temperatures:
Sea temperatures also rose in the northern North Atlantic (Fig. 4). Jensen (1939) documented the increase in sea surface temperatures (SSTs) off West Greenland while Scherhag (1937) and Smed (1947, 1949) did the same for several areas of the northern North Atlantic. Smed’s (1947) analysis showed the largest increase was off West Greenland and Norway, although no data from north of the Arctic Circle were analyzed (see also Beverton and Lee, 1965). A significant but lower sea surface temperature increase was observed off southwest Iceland (Fig. 4; Thomsen, 1937). The increase at the Faroes was of the order 0.5 C and occurred in the early 1930s (Ta˚ning, 1953), somewhat later than many other locations but consistent with the later rise in air temperatures observed at this site (Fig. 2). Sea temperatures also rose farther south in the North Sea, the English Channel and the Baltic Sea (Beverton and Lee, 1965). Scherhag (1937) reported that Gulf Stream temperatures were approximately 0.4 C higher in 1926–1933 compared to 1912–1918.
And it was not just sea surface temperatures:
The rise in ocean temperatures was not restricted to the surface waters. Jensen (1939) and Dunbar (1946) documented subsurface temperature increases down to 500 m in the waters off West Greenland. On the other side of the Atlantic, the annual average temperature of the top 200 m in the Kola section in the Barents Sea (Fig. 1) shows a sharp rise during the 1920s of over 0.5 C and remained high through to the early 1960s before declining (Fig. 5).
The fact that subsurface temperatures were rising in Arctic waters is hugely significant, as it shows it as not just a surface event. Once warm waters enter the Arctic basin, they tend to stay a long time, just circulating around the gyres. This is evidenced by the fact that Arctic temperatures remained high until the early 1960s.
And, unsurprisingly, all of this had a marked effect on sea ice:
There was a marked reduction in sea-ice extent that accompanied the warming. Polar ice or ‘‘storis’’ commonly flows from East Greenland around Cape Farewell and along West Greenland. The northernmost position of the storis off West Greenland, by month, and the number of months storis appeared between October and the following September, for the years 1899/1900 to 1971/1972 were listed by Valeur (1976). May is typically the month with the maximum northerly extent of storis. During the period 1899/1900 to 1925/1926 the ice penetrated farther north during May and lasted almost 2 months longer than for the period 1926/1927 to 1960/1961 (Fig. 6). Off Iceland, in the 1920s and 1930s, the ice edge retreated northward such that ice was observed along the north coast for only a couple of weeks each year compared to 12–14 weeks in the late 1800s (Lamb and Johnson, 1959; Schell, 1961). In the Barents Sea, the ice edge moved north and east (Beverton and Lee, 1965). Around Svalbard it retreated an average distance of some 150 miles northward in the late summer (Brooks, 1938). These retractions began in the Barents Sea and the Kara Sea in 1920 (Ahlmann, 1949).
Drinkwater is not the only scientists to describe this dramatic Arctic warming as the most significant climatic event in the 20thC. As we know, current temperatures in most parts of the Arctic are no higher than they were back in the 1930s and 40s, and we see the same pattern of warmer waters flowing into the Arctic.
The reality is that there is nothing unprecedented about current warming, and no reason why it should be caused by CO2. The fact that the rapid cooling that followed in the 1960s and 70s occurred at a time when CO2 emissions were rising rapidly is surely evidence of this.
Whither next for the Arctic? I have not got a crystal ball, but it would be a brave man who refused to countenance the possibility that the cold cycle will return at some stage in the future.