Andrew Russell: weather and climate scientist

Southern Hemisphere Atmospheric Circulation Impacts on Eastern Antarctic Peninsula Precipitation

Supervisors: Prof. Glenn R. McGregor (Now at University of Auckland)
Dr. Gareth J. Marshall (British Antarctic Survey)

"This is an interesting and important study and I would qualify the scientific level of the study as very good"
Dr. Michiel van den Broeke (External Examiner)

"I found the thesis most systematic and thorough ... the way in which the author utilizes climatological, meteorological and ice chemical data is sophisticated and convincing"
Prof. Ian Fairchild (Internal Examiner)

Thesis Abstract

The nature of precipitation delivery mechanisms for Dolleman Island (DI) eastern Antarctic Peninsula (AP) was investigated by undertaking a climatological analysis of air mass back trajectory (BT) patterns stratified by three precipitation thresholds as defined using ECMWF 40-year re-analysis data. This revealed the principal precipitation sources for DI to be: (1) the Weddell Sea, with blocking west of the AP; (2) the South Atlantic when the Circumpolar Trough (CPT) is weak; and (3) the South Pacific when the CPT is deep. The intensity of the CPT, the phase of the Southern Annular Mode (SAM) and the occurrence of ENSO events also influence precipitation over DI.

BT patterns were used to interpret the inter-annual variation in ice core chemistry for DI. Nitrate concentration was found to be strongly correlated with the ratio of easterly to westerly BTs, as described by a Cross-Peninsula Index (CPI). The CPI was subsequently used to reconstruct general atmospheric circulation characteristics for the 340-years represented by the DI ice core. This confirmed the importance of the westerlies but highlighted a period of increased easterlies during 1720-1780 and increasing westerlies over the period 1950-1980, concomitant with a predominantly positive phase of the SAM. The reconstruction also revealed periods when polynyas may have dominated the Weddell Sea and confirms the occurrence of historical El Niņo events as found in other palaeoclimatic reconstructions.

Relevant Papers

I've written a couple of papers from this work, see my publications page for pdf versions of this work.

Summary for Website

The Antarctic Peninsula is one of the "hot-spots" of climate research. Over the last 50 years, it has undergone one of the largest temperature increases on the planet and it is hypothesysed that changes in atmospheric circulation are a significant contributor to this temperature change. This undoubtedly had an influence on the collapse of the Larsen B ice shelf in 2002. However, the Antarctic has only very short records of observations for the major meteorological fields. If we wish to gain further perspective on the changes occurring in this region we have to reconstruct characteristics of atmospheric circulation by understanding the links between present day circulation patterns and the signals that are recorded in ice core records from the Antarctic. This was the aim of my PhD project.

The first step in this process is to understand the current precipitation delivery mechanisms. This figure shows composite plots of clusters of 3-D atmospheric back trajectories that were run for days when precipitation fell at Dolleman Island (an ice core site on the eastern Antarctic Peninsula). This was done for the period 1979-1992 as this represents the overlap period of reliable trajectories and ice core data. The important result here is that there is a clear distinction between the trajectories that approach the site from the east and the west. The next step in this process is to try and link these precipitation delivery mechanisms with signals in the ice core.
It turned out that there were strong positive correlations between the chloride and nitrate ice core measurements and the frequency of westerly trajectories. This correlation was reversed for the frequency of easterly trajectories. This figure summarises the links between the atmospheric circulation and the ice core signals. There are several factors at work that influence what is eventually detected in the ice core. The atmospheric factors here are mainly the phase and strength of ENSO and the SAM - these determine where the air arriving at the site is coming from. This is not the whole story, though, as other factor such as sea ice extent, polynyas and volcanic eruptions have a large effect on the ice core signals.

Knowing all this allows us to confidently produce a regression equation which will give an indication of nature of the precipitation delivery mechanisms from the ice core. The easiest way of doing this is to use an annual ratio of westerlies to easterlies arriving at the site and regressing this onto the ice core data. Once this has been done for the period of overlap between the ice core data and the back trajectories, the regression equation can be applied to the full length of the ice core. The results of this process can be seen in this figure. This gives a hitherto unknown indication of the atmospheric circulation characteristics in this region, which is a reasonably interesting result! In particular, as mentioned in the abstract, this analysis highlighted a period of increased easterlies during 1720-1780 and an increase in westerlies over the period 1950-1980, concomitant with a positive trend in the SAM and western AP warming. The reconstruction also revealed periods when polynyas may have been present over the Weddell Sea.