Search Results

You are looking at 1 - 10 of 15 items for :

  • Author or Editor: Neville Nicholls x
  • Refine by Access: All Content x
Clear All Modify Search
Neville Nicholls

Abstract

Canonical correlation is proposed as an exploratory technique for studying teleconnections. It is suggested that the technique can elucidate the temporal signature (i.e., the seasonally varying nature) of the teleconnections and the lags between variables. Several teleconnections studied with the Southern Oscillation are subjected to canonical correlation as examples. The teleconnections studied are between Darwin pressure and Tahiti pressure, southeast Australian rainfall, and Willis Island air temperature. In each example the canonical correlation analysis confirms the teleconnections uncovered previously by other statistical techniques but also suggests the existence of other interesting features of these teleconnections.

Full access
Neville Nicholls

Abstract

The number of tropical cyclones observed in the Australian region during a single cyclone season has ranged from one to nineteen since 1909. Previous studies, using limited data sets, have suggested that interannual variations in the number of cyclones are related to the Southern Oscillation and that an index of the Southern Oscillation (e.g., Darwin pressure) can be used to predict the number of cyclones expected in the coming season. This study uses a 74 year time series of tropical cyclone numbers, from the 1909/10 season to the 1982/83 season to confirm this. Strong and stable correlations are found between cyclone numbers and Darwin pressures before and during the cyclone season. Even stronger relationships are found between Darwin pressure and the number of cyclone days in a cyclone season. The correlations are strong and stable enough to allow prediction of seasonal cyclone activity from several months prior to the start of the tropical cyclone season. A simple equation for predicting seasonal cyclone activity is derived.

Full access
Neville Nicholls

Abstract

Evidence is presented supporting the hypothesis (first expressed over 60 years ago) that interannual fluctuations of early wet season rainfall in the Indonesian Archipelago can be successfully predicted from prior observations of atmospheric pressure anomalies. It is shown that this predictability is related to sea surface temperature anomalies. The postulated mechanism for this predictability is interaction of the atmosphere and ocean leading to a tendency for anomalies in the two media to persist. Experiments to testthis postulate are suggested.

Full access
Neville Nicholls

Abstract

The mouths of October 1971 and 1972 were, respectively, months of heavy and light rainfall over much of Indonesia and New Guinea. The two months also showed considerable differences in the atmospheric circulation in the higher latitudes of the Australian region, with the wetter month (1971) exhibiting stronger westerlies between 30° and 45°S and weaker tropospheric westerlies poleward of 45°S. These changes are similar to those observed in the eastern Pacific during periods of heavy or light equatorial rainfall. Examination of 14 years of data in the Australian region suggests the presence of 1) significant coherence between tropical rainfall and the strength of the subtropical westerlies for periodicities greater than six months with no apparent lag, and 2) a significant out-of-phase relationship between tropical rainfall and high-latitude westerlies at periods greater than 24 months, the latter lagging the former by several months. The relevance of these results to tropical-extratropical interactions and numerical modeling is discussed.

Full access
Neville Nicholls

Abstract

A rotated principal component analysis of Australian winter (June–August) rainfall revealed two large-scale patterns of variation which together accounted for more than half of the total rainfall variance. The first pattern was a broadband stretching from the northwest to the southeast corners of the country. The second was centered in the eastern third of the continent. The two patterns were correlated to sea surface temperatures in the Indian and Pacific oceans. The first rainfall pattern was best related to the difference in sea temperatures between the Indonesian region and the central Indian Ocean. The second rainfall pattern was related to equatorial Pacific sea surface temperatures. This relationship reflects the influence of the Southern Oscillation on both sea surface temperatures and Australian rainfall but the relationship between the first rainfall pattern and the difference between Indonesian and central Indian Ocean sea surface temperatures is largely independent of the Southern Oscillation. This sea surface temperature difference may be another factor influencing Australian rainfall, some-what separate from the well-known effect of the Southern Oscillation.

Full access
Neville Nicholls

A number of studies in meteorological journals have documented some of the constraints to the effective use of climate forecasts. One major constraint, the considerable difficulty people have in estimating and dealing with probabilities, risk, and uncertainty, has received relatively little attention in the climate field. Some of this difficulty arises from problems known as cognitive illusions or biases. These illusions, and ways to avoid them impacting on decision making, have been studied in the fields of law, medicine, and business. The relevance of some of these illusions to climate prediction is discussed here. The optimal use of climate predictions requires providers of forecasts to understand these difficulties and to make adjustments for them in the way forecasts are prepared and disseminated.

Full access
Neville Nicholls
and
Alex Kariko

Abstract

The number, average length, and average intensity of rain events at five stations located in eastern Australia have been calculated for each year from 1910 to 1988, using daily rainfall totals. A rain event has been defined as a period of consecutive days on which rainfall has been recorded on each day. Inter-relationships between the rain-event variables (at each station and between stations), along with their relationships with annual rainfall and the El Niño-Southern Oscillation, have been investigated. Trends in the time series of the rain-event variables have also been examined. Annual rainfall variations are found to be primarily caused by variations in intensity. Fluctuations in the three rain-event variables are essentially independent of each other. This is due, in some cases, to inter-relationships at interdecadal time scales offsetting relationships of the opposite sense at shorter time scales. The large-scale geographical nature of east Australian rainfall fluctuations mainly reflects interstation correlations in the number of events. The El Niño-Southern Oscillation affects rainfall mainly by influencing the number and intensity of rain events. Twentieth century increases in east Australian rainfall have been due, primarily, to increased numbers of events. Intensity of rain events has generally declined, offsetting some of the increase in rainfall expected from more frequent events.

Full access
Andrew Solow
and
Neville Nicholls

Abstract

A statistical model of the relationship between tropical cyclone frequency in the Australian region and an index of the strength and phase of the southern oscillation is developed for the period 1910–88. The modeling is nonstandard because the cyclone record is incomplete early in this period. The fitted model indicates that the mean annual number of cyclones during a major cold event is twice that during a major warm event.

Full access
Roger Stone
,
Graeme Hammer
, and
Neville Nicholls

Abstract

A forecast method capable of estimating date of last frost and number of frosts per season in northeastern Australia some months in advance is described. Forecast “skill” is achieved using either Southern Oscillation index (SOI) patterns (phases) during the previous austral autumn or a linear discriminant approach and the SOI. When applying these systems, it is possible to provide significantly different probability distributions of day of last frost and number of frosts, depending on the SOI patterns observed during the previous season. An analysis of the time series of frost frequency and date of last frost suggests an apparent warming trend in the data, resulting in a trend toward earlier dates of last frost and fewer numbers of frosts at many of the locations analyzed. The beneficial implications of the proposed frost forecasting system for enterprises such as winter agriculture in the region are believed to be significant.

Full access
Scott Power
,
Brian Sadler
, and
Neville Nicholls

Water flow into dams that supply Perth in Western Australia (WA) has fallen by 50% since the mid-1970s, and this has severely tested water managers. Climate change scenarios available since the 1980s have suggested that global warming will reduce rainfall over southern Australia, including Perth. Water managers recognize the uncertainties associated with the projections, including the significant differences that exist between the timing and magnitude of the observed changes and modeled projections. The information has, nevertheless, influenced their decision making.

To understand why, we need to consider the broader environment in which the water managers operate. One key factor is that the imposition of severe water restrictions can lead to significant economic loss and increased unemployment. Prolonged restrictions can therefore create strong debate in the wider community. In recognition of this, state government policy requires that water managers ensure that the chance of having severe restrictions is kept low. Severe restrictions have not been imposed since 1979, but moderate restrictions are more common, and were imposed as recently as 2002. Scrutiny of water management can become intense even after moderate restrictions are imposed, and at these times it is unacceptable to many people if a known risk—even if very uncertain—is perceived to have been ignored in earlier planning. Climate science has established regional drying driven by global warming as a risk, and so global warming has to be addressed in planning. Water managers also need climate science to reassure the public that the restrictions imposed were necessary because of unprecedented changes in rainfall, not because of poor management.

In recent years much of the influence that climate science has had on water managers can be attributed to the Indian Ocean Climate Initiative (IOCI). IOCI is a research partnership between the Western Australia Water Corporation, other state government agencies, and two national meteorological research organizations. Water managers saw their participation in IOCI as one strand of a broader risk management plan. They did not have the luxury of deferring important decision making for certainty that climate science might never bring, but were very interested in what climate science might provide “now.”

The participation of water managers in IOCI enabled them to influence research planning to better meet their needs. Water managers did not just want predictions or technical explanations of an individual scientist's latest work. They wanted reliable and balanced advice on broader issues, explanations, clarification, realistic expectations, and an appreciation of uncertainty. They wanted climate information related to water management issues in a form relevant to the region. “Localized” information is more suitable for inclusion in their decision making, and of more use to them for both informing, and stimulating discussion within, the wider community.

Full access