An Interview with Robert Hilton

Abstract

Calibration of proxy climate records is well-established for annually resolved proxies such as tree rings, but it has not been attempted for non-annually resolved proxies such as those from peatland surface wetness records. Several previous studies have suggested that peatland surface wetness is primarily driven by warm season moisture balance and implied a potential for producing calibrated records of deficit or precipitation. This paper presents a high-resolution testate amoebae analysis of a peat record from central Ireland covering the last c.1000 years, and provides the first attempt to produce a calibrated record of past precipitation from a peat record. Past water table depth was reconstructed using a transfer function applied to contiguous samples for the top 1 m of the profile. The chronology was constrained by a series of radiocarbon ages (including ‘bomb-spike’ ages) and spheroidal carbonaceous particles. Correlations between reconstructed water tables and meteorological records (1958–1995) of precipitation and deficit were strongly positive and were used to reconstruct precipitation and deficit from the last 1000 years. Validation using earlier meteorological records was hampered by very low peat accumulation rates, but suggested the summer precipitation and deficit reconstructions were more robust than a reconstruction of annual precipitation. The summer precipitation reconstruction suggests that the period AD1400–1850 experienced higher summer rainfall than for much of both the preceding 400 years and the last 150 years. The change in regime from low to high summer rainfall may be attributed to changes in the summer North Atlantic Oscillation. Combined with tree ring and speleothem records of winter NAO, this suggests a major change in seasonality of precipitation in far western Europe between the Medieval Climate Anomaly and the Little Ice Age. The MCA was characterised by dry summers and wet winters, whilst the LIA had wet summers and dry winters. Calibration of peat surface wetness records using meteorological records holds much potential for the future and may lead to improved insights into seasonal precipitation and water balance changes. This study was limited by slow accumulation rates leading to low temporal resolution for the late 19th and early 20th century part of the record. Further development of the technique will require more highly temporally resolved records of change over the whole of the instrumental time period to allow a full calibration and validation approach to be applied.