Indicator: Rainfall & its Variability

| Indicator description | Complete list of Indicators |

What the results tell us for Tumut

Analysis, graphs and tables by

Rainfall statistics

Tumut was drier than normal in 1997 and 1998, with 287 mm less rainfall than normal (an anomaly of 34%) in 1997 (annual rainfall totals are compared with the long-term mean in the table below).

In 1999 rainfall was close to average. The extended period of below-normal rainfall related to the El Niņo event of 1997/1998 is evident at Tumut, as is the wetter than normal winter of 1998 (1997-1999 monthly rainfall graphs below).

It is informative to consider these rainfall statistics in the context of statistics for a range of climatic parameters in Tumut, and in the context of longer-term rainfall patterns, and the climate zones we have identified in this Region.

Tumut was drier than normal in 1997 and 1998, with 1997 the drier year (34% below-average rainfall). In 1999 annual rainfall was slightly above average. The graphs below show the pattern of monthly rainfall over the three-year period, as both monthly rainfall totals (upper graph) and monthly rainfall anomalies (lower graph).

Figure 1. Monthly rainfall totals (in mm) in Tumut. The annual rainfall total is the sum of the monthly totals. If a month has missing data the month is classed as missing, and no data are shown. Figure 2. Deviations from mean monthly rainfall (in mm) in Tumut. These are calculated as the difference between the rainfall for each month and the mean rainfall for that month throughout the record.

Ten months had below-average rainfall in 1997, particularly April and October. Only May and September were wetter than average. Similarly, in 1998 seven months had below-average rainfall (March and May were particularly dry). In 1999 six months were drier than normal, with July having the largest negative anomaly (37% of long-term mean rainfall). The longest unbroken period of below-average rainfall extended from October 1997 to March 1998. The winter/spring of 1998 (June-September) was consistently wetter than average. In absolute terms, the driest month in the three-year period was March 1998, with 1 mm (98% below normal for that month); the wettest month was December 1999, with 130 mm (127% above the monthly average of 57 mm).

The climatological context for rainfall in Tumut

The long-term perspective

As is common in the region, there is a high degree of interannual rainfall variability at Tumut. The range of variability has changed little over time, however, although conditions were slightly drier than normal early in the record and somewhat wetter in the second half. These features are evident in the graphs of annual rainfall at Tumut, below, which provide a long-term perspective on rainfall variability over the last 114 years.

Figure 3. Tumut: annual rainfall totals (in mm) span the full length of the station record (1886-1999). The long-term mean rainfall (horizontal line) provides a reference for comparison of individual annual totals. Figure 4. Tumut: annual rainfall from 1886 to 1999 showing deviations from the median, calculated as the difference between the rainfall for each year and the mean for all years (scale in mm). A four-year running mean indicates the underlying pattern of interannual variability.

Following a prolonged period of below-average rainfall (about 20 years) around the turn of the century, rainfall at Tumut has gone through a series of shorter wet and dry spells of varying duration. The wettest period was that from about 1950 to the mid 1970s, although there was a high degree of interannual variability during this time. More recently, Tumut was drier than normal for approximately six years around 1980; thereafter conditions became slightly wetter. The wettest year on record was 1956 (1308 mm); the driest was 1967 (340 mm).

As has occurred throughout the region, rainfall at Tumut has undergone clear fluctuations in seasonality since the record began in 1886. On average August is the wettest month and February the driest, with a range of more than 40 mm.

Figure 5. Seasonal variation in monthly rainfall: Tumut Click to view larger version A smoothed view of the variation of monthly rainfall (y-axis: January at bottom, December at top) through time (x-axis) over the last 114 years. The data are expressed as anomalies (i.e. the difference between the individual month's rainfall and the long-term average for all months in the data set). The colour scale indicates the intensity of the rainfall anomaly, ranging from white and brown (very dry), through yellow (slightly dry), to green (slightly wet) and dark blue (very wet). Contours are at 10 mm intervals. Coherent patterns of variation through time are evident in this type of plot; e.g. some months wetter early this century, then becoming drier. Another common example is of a shift in the timing of the wettest or driest part of the year, which can have important consequences for agriculture. The data have been smoothed using a non-linear 61-month filter; this means that variations at shorter time scales than about 5 years are smoothed over, revealing lower-frequency patterns of change in seasonality.

The drier and wetter seasons at Tumut have remained approximately stable over time, with dry conditions in summer and autumn (November-May) and wetter conditions in winter and spring (June-October). The exception is the intrusion of drier conditions into June after 1940. The principal changes have been in the intensity of the seasonal cycle. The drier season has been centred on February throughout the record, but with most intense dry periods around 1900 and 1960. Between 1900 and 1920, the wetter season was also particularly well defined in June; there was a difference in rainfall of more than 60 mm between these two months. Since 1980 the wettest month has been July, and there has been a difference in rainfall of approximately 40 mm between February and July. Overall June rainfall has decreased (as it has done elsewhere in the region), but other months show smaller changes than are typical at many other stations.

The underlying trends in the record of rainfall at Tumut help to clarify patterns in the apparently random interannual fluctuations. It should be remembered that particularly high or low values at the start or end of a record can have a disproportionate effect on the shape of the curves.

Figure 6. Low-frequency fluctuations in monthly rainfall: Tumut The graph shows the underlying characteristics or trends in the monthly rainfall (y-axis, in mm) over time. Three different smoothers have been used: a low-frequency one (~40 years), that shows the lowest variability; a medium-frequency one (~20 years) that shows moderate variability; and a higher-frequency one (~10 years) showing the greatest variability. When looked at together, these filtered rainfall series reveal interesting characteristics of the behaviour of rainfall over time. Long-term trends are evident, as are periods when rainfall variability was generally greater or less, and periods when fluctuations of different frequencies were dominant.

Tumut rainfall shows a decline in the long-term trend from the start of the record to about 1910, followed by a general rising trend to the late 1950s. Since then rainfall has tended to decline until around 1980, and has subsequently increased slightly. There have been periods of complex interaction between the medium- and shorter-term components of rainfall variation, interspersed with shorter periods when the two have fluctuated in phase and have reinforced the tendency to a wetter or drier than normal spell of years. Examples of the latter occurred in the late 1920s and around 1980.

As is true throughout most of south-eastern Australia, there is a relatively weak direct relationship between the Southern Oscillation Index (SOI), a measure of the behaviour of the ENSO phenomenon, and rainfall at Tumut. This means that higher rainfall tends to occur when the SOI is strongly positive (e.g. 1955), and lower rainfall when it is strongly negative (e.g. 1982). This is illustrated in the graphs below, which show the degree of association between annual rainfall totals and annual average values of the SOI. However, the association is not strong (the correlation coefficient is 0.40) and some very wet and dry years have occurred with the SOI close to zero. The large-scale atmospheric fluctuations measured by the SOI account for only around 15% of interannual rainfall variability at Tumut.

Figure 7. The scatter plot indicates the correlation between station rainfall the SOI; a tight, generally linear grouping of points suggests a good correlation, whereas a wide, random scatter indicates a poor relationship. Figure 8. The time series highlights variations in the relationship between the SOI and rainfall over time. It shows the deviations from mean annual rainfall as vertical bars, with the SOI plotted as a continuous line. SOI values have been multiplied by 10, for scaling reasons.

The July-June 12-month period was used instead of the calendar year in this analysis, because of the characteristics of the SOI and of its association with summer rainfall (September- March) in south-eastern Australia.

About the data

See also: Information about the analysis techniques used

Rainfall has been measured at a total of 47 official Bureau of Meteorology stations in Tumut Shire, 15 of which are currently operational. One of these is the long-term station at Tumut. Other organisations (e.g. the Snowy Mountains Authority or NSW Parks and Wildlife) may make independent measurements relevant to the area; their data may be made available independently, or through the Bureau of Meteorology. There are also likely to be a number of privately-kept records in the Shire; these often contain invaluable information for otherwise sparsely-monitored areas.

The Tumut rainfall record commences in 1886, and is suitable for long-term climatological analysis. Long rainfall records from stations close to Tumut Shire (in Yass, Yarrowlumla or Snowy River shires, for example) may also provide insights into the rainfall climatology of the Shire.

Information on Bureau of Meteorology weather stations and climate data may be obtained by contacting the National Climate Centre in Melbourne:
Tel: (03) 9669 4082
Fax: (03) 9669 4515
Email: dstran@bom.gov.au
http://www.bom.gov.au

Full-page view of Figure 5. Seasonal variation in monthly rainfall: Tumut
Back to Figure 5.

A smoothed view of the variation of monthly rainfall (y-axis: January at bottom, December at top) through time (x-axis) over the last 114 years. The data are expressed as anomalies (i.e. the difference between the individual month's rainfall and the long-term average for all months in the data set). The colour scale indicates the intensity of the rainfall anomaly, ranging from white and brown (very dry), through yellow (slightly dry), to green (slightly wet) and dark blue (very wet). Contours are at 10 mm intervals.

The data have been smoothed using a non-linear 61-month filter; this means that variations at shorter time scales than about 5 years are smoothed over, revealing lower-frequency patterns of change in seasonality.

Description: What does 'rainfall and its variability' measure?

Which data are collected?

• rainfall statistics
• range of variability

Why do we report this indicator?

Of all the standard climatic parameters, rainfall is the most variable in time and space. The amount of rainfall received across the Australian Capital Region can vary dramatically from year to year, ranging from dry periods that can persist for years, to periods of intense downpours, storms and flooding. On average, the region falls into the zone of uniform seasonal distribution of rainfall (i.e. it has no clearly defined wet or dry season), but this too can vary greatly between years.

Together with temperature, the features of rainfall variability in time and space across the region affect the distribution of the flora and fauna. Many species have adapted to cope with the wide range of variability that has characterised Australian rainfall for millennia. Rainfall is also a key determinant of the growing season and the types of agriculture practised - as water is limiting for many human activities. It directly influences surface runoff, streamflow and groundwater recharge.

There are several measures of rainfall that can be used to provide information on variability and hence on the state of the climate. These include:

• annual rainfall totals
• rainfall seasonality
• the frequency of extreme events.

Data for each of these are considered for the current reporting period, and for the longest period for which data are available; wherever possible they are compared with long-term mean values.