Centennial climate variability in the British Isles during the mid-late Holocene

Abstract

Multi-millennial climate changes were relatively minor over the mid–late Holocene in the British Isles, because orbitally forced insolation changes were smaller than those at higher latitudes. Centennial climate variability is thus likely to have exerted a greater influence on the environment and human society of the region. Proxy-climate records from the British Isles covering the last 4500 years are assembled and re-evaluated with the aim of identifying centennial climate variability reflected by multi-proxy indicators. The proxies include bog oak populations, peatland surface wetness, flooding episodes from fluvial deposits, speleothem annual band width and oxygen isotopes, chironomids from lake sediments and sand and dune deposition. Most proxies reflect water balance rather than temperature alone, and records predominantly reflect warm season climate. A series of 12 key periods of enhanced precipitation–evaporation (P-E) are identified by their presence in two or more proxy records. Variability in P-E is much greater than that shown by temperature proxies and there is no necessary association between warm/cool and dry/wet periods. Although the data for temperature are less robust than those for P-E, a series of key temperature changes are proposed based on speleothem δ18O and chironomid inferred July temperature records; relatively cool before c. 3100 years BP, warmer (3100–2000 years BP), cool (2000–1250 cal years BP), warm (1250–650 cal years BP), and cool (650 cal years BP onwards). Some key increases in P-E (2750, 1650, 550 cal years BP) show a strong correspondence with ‘Bond cycles’ in ocean proxy records for increased ice rafted debris, decreased summer sea surface temperatures and sometimes decreased North Atlantic deep water circulation. Other higher frequency changes in P-E are also strongly related to SST variability. Whilst some of the main changes to cooler SSTs and increased P-E are approximately coincident with reduced solar output, most are not and thus must be the result of the internal dynamics of the ocean and atmosphere. Future work should concentrate on firmly establishing the pattern of temperature change, improving chronological accuracy and precision in existing records and improving process-based understanding of proxies.