Transcript Revision of Intensity Frequency Duration Data: Pilot

The effect of changes in dew-point temperatures on
estimates of Probable Maximum Precipitation
R. J. Smalley, D. Jakob, D. A. Jones, J. Meighen, B. F. Taylor, K. C. Xuereb
Bureau of Meteorology Research Centre, Bureau of Meteorology
GPO Box 1289, Melbourne VIC 3001
[email protected]
1. Introduction
4. Estimates of storm dew-points for Brisbane
The Australian Bureau of Meteorology provides methods and data for estimating
Probable Maximum Precipitation (PMP). These are used in the design of dam
spillways, determination of the existing flood capacity or for floodplain
management.
Figure 2 shows the storm dew-points obtained using the different estimates for each
significant rainfall event. Also shown is the weighted mean from three estimates
[Tdm=((Tde1 + Tde2)/2 +24hTdh)/2]. Overall, the data indicate a lack of agreement
between the estimates. Relative to Tdm, the bias of 24Tdh and its 95% confidence
interval are 0.4°C and ±2.5°C respectively. The standard error is of the order ±0.5°C.
Definition of Probable Maximum Precipitation (PMP)
22.0
PMP is defined as (WMO, 1986) ‘The greatest depth of precipitation for a
• over a given size storm area
20.0
’
Storm dew-point (°C)
given duration meteorologically possible
• at a particular location
• at a particular time of the year,
’
18.0
16.0
Tdh2
Tdh
Tde1
Tde2
Tdn
Tdm
14.0
12.0
10.0
• with no allowance made for [future] long term climatic trends.’
8.0
Recent results (e.g. Abbs 2004) using climate model output, have indicated for
Australia changes in the spatial distribution of rainfall intensities in a changing
climate. This indicates that the potential effects of climate change on PMP
estimates should be assessed.
3
5
3
8
8
3
6
9
6
l 08
r1
r0
r2
v0
y0
y0
n1
n2
n2
Ju
Ja
Ja
Ja
Ma
Ap
Ap
No
Ma
Ma
73
68
71
74
74
89
89
81
80
96
19
19
19
19
19
19
19
19
19
19
Event nam e
Figure 2. Comparison between estimates for the storm dew-point temperatures
2. Methods of PMP estimation
5. Trends in storm dew-points
The PMP is a theoretical quantity and cannot be directly measured.
Depending on the location and duration, one of three generalised
methods is used for estimating the PMP:
Means and frequency distributions of the storm dew-points across the periods 19571980 and 1981-2003 were calculated. Only the 6 months corresponding to the
selected significant rainfall events (from Figure 2) are shown in Figure 3. Also
indicated are the storm dew-points for each significant rainfall event and the mean
for each time period. Only one event (26 April 1989) is within the positive tail of its
associated distribution. For the storm dew-points, May is the only month to exhibit a
difference in the means (1.6 °C) between the two periods.
• Generalised Tropical Storm Method-Revised (GTSMR, is used for areas
where extreme events are likely to be caused by tropical rainfall events)
• Generalised South-East Australia Method (GSAM)
• Generalised Short Duration Method (GSDM, for durations below 6 hours)
0.2
January
March
April
0.15
The generalised methods (with
their applicable areas shown in
Figure 1) maximise a number of
physical parameters based on
previous extreme rainfall events.
These include
0.1
0.05
1981-2003
24hTdh
mean
1957-1980
24hTdh
mean
0
0.15
May
• moisture availability and
November
June
0.1
• storm efficiency
0.05
The current study will focus on
these factors.
0
Figure 1.
The generalised method areas
0
10
24hTdh (°C)
20
0
10
24hTdh (°C)
20
0
10
24hTdh (°C)
20
Figure 3. Distributions of 24T dh for two time periods (1957-1980 and 1981-2003)
6. Conclusions and further work
3. Moisture Availability
The storm precipitable water (SPW) in a column of air is used as a measure of the
available moisture for a particular rainfall event (or ‘storm’). In the generalised
methods used for estimating the PMP, at a particular location and time of year, the
moisture availability is maximised (providing the extreme precipitable water, EPW).
Since an estimate of the PW determined from upper air data is not usually available,
the 24h persisting surface dew-point temperature is used as a surrogate. Dew-point
temperature data from 57 stations across Australia have been homogenised, which
ensures that non-climate related in-homogeneities are identified and corrected.
Estimates of SPW and EPW can be obtained from the homogenised 24h persisting
dew-point data (24hTdh), by assuming a saturated atmosphere, with a pseudoadiabatic lapse rate. To determine the accuracy of using 24h persisting dew-point
data, comparison with estimates from upper-air data (Tde1 and Tde2) are initially
carried out at Brisbane for 10 significant rainfall events between 1957 and 2003.
Five different estimates are compared. These are described in Table 1.
Table 1. Notation used for estimates of the storm dew-point temperatures
All
observations
0900 & 1500
observations
Nearby
stations
Integrated
upper-air
Extrapolated
upper- air
24hTdh
24hTdh2
24hTdn
Tde1
Tde2
An analysis of storm dew-point temperatures, using all available dew-point
estimates was carried out for 10 significant rainfall events at Brisbane between
1957 and 2003. The most reliable estimate is obtained using the 24hTdh at the
station of interest. This has a typical error of ±0.5 °C, with an error range at the
95% level of -2.1°C to 2.9°C. To better quantify the possible influence of climate
change on moisture availability and PMP estimates, this study will be expanded by:
• Identifying 20 significant rainfall events for each of the two intermediate periods
at each selected station
• Using these events to test for significant changes in the storm dew-point
temperatures
Acknowledgement
This project is funded jointly by the Australian Greenhouse Office (AGO), the
Queensland Department of Natural Resources, Mines and Water (NRMW) with inkind contributions by the Bureau of Meteorology. Dr C. Lucas is gratefully
acknowledged for developing and providing the high-quality dew-point data.
References
Abbs, D. J., (2004) The effect of climate change on the intensity of extreme rainfall
events presented at International Conference on Storms. AMOS 11th National
Conference, Brisbane, Australia, 5-9 July.
World Meteorological Organisation (1986). Manual for Estimation of Probable
Maximum Precipitation. Second Edition. Operational Hydrology Report No. 1, WMO
– No. 332, Geneva.