Home | About Us | Contacts | Help | Feedback | SEARCH Global | Australia | NSW | Vic. | Qld | WA | SA | Tas. | ACT | NT | Ant. | Weather & Warnings | Hydrology | Climate | Numerical Prediction | About Services | Learn About Meteorology | Registered User Services |

Please note that we are currently experiencing some delays in the delivery of products and services to our website of up to 45 minutes at times.

An improved system for producing objective daily, weekly and monthly national and regional scale analyses of rainfall has been developed through a collaborative effort involving several units of the Bureau of Meteorology. Daily rainfall charts from this system are routinely produced in the National Meteorological Operations Centre. These analyses are available at approximately 1:30pm Eastern Standard Time, prior to an update at 3:30pm to allow for occasional additional data. If the This page shows the full national analysis as well as a high resolution analysis for SE mainland Australia. At present, an early analysis is available for all states except WA at approximately 1:30 pm Eastern Standard Time. To see these charts, go to the archive of images , and view 'early_rain_latest.gif'. The full national rainfall analyses may be available earlier in the archive of images than on this page - view 'rain_latest.gif'. The SE Australian analysis may be found in 'latest_serain.gif'.

A 3-pass Barnes-type successive correction method is used to prepare the operational rainfall analyses. Analyses can be further refined and extended into no-data areas through the application of satellite "no rain" pseudo-observations and "Indicator Kriging" (see Experimental Products ). Statistical interpolation has also been investigated as an alternative analysis scheme, along with generalised cross-validation to further improve observation quality control (again, see Experimental Products ).

Contacts: Gary Weymouth, Beth Ebert, Sun Xudong

Authors of operational rainfall analysis system: G.A.Mills, G.Weymouth, D. Jones, E.Ebert, J.Lorkin, and J.Kelly.

Archive of images

Archive of data

Archive of analysis grids

Mills, G.A., G. Weymouth, D. Jones, E.E. Ebert, M. Manton, J. Lorkin and J. Kelly, 1997: A national objective daily rainfall analysis system. BMRC Techniques Development Report No. 1, Bureau of Meteorology Research Centre, Melbourne, 30 pp.

Several experimental rainfall analysis products are currently under development. The first is an analysis performed using Statistical Interpolation (SI). Statistical interpolation offers significant advantages over successive corrections schemes (such as the Barnes scheme) in that directional effects can be properly accounted for. It also facilitates the inclusion of independent observations from other sources such as radar and/or satellite. Both a national analysis and a higher resolution southeastern analysis are produced daily.

Kriging is a stochastic linear interpolation method that estimates the rainfall by minimising the estimated error variances of a regionalised variable. Indicator Kriging (IK) is a statistically optimised method of delineating the raining area in particular. To implement IK, a binary variable is used to describe the rainfall from rain gauge network. Meanwhile, satellite rainfall probability of occurrence has been calculated from its infrared temperature. Kriging is used to separately analyse the indicators from the rain gauge and satellite data. The two indicator analyses are then merged using optimal weighting. Finally, the gauge observations are interpolated (again using kriging) only the regions determined to be raining. An advantage of this approach is that satellite data can help delineate the non-raining areas, thus improving the analysed rain extent and helping to reduce the data gap in gauge-sparse regions.

Experimental analyses can be viewed by selecting items from the table below. The date in the analysis refers to the date of the 9 a.m. local time reading, which corresponds approximately to 00-24 UTC of the previous (valid) day.

Valid Date, Time	Operational Barnes	Experimental SI	Experimental SI (SE)	Indicator Kriging	Rain Kriging
20020313 00-24 UTC	rain_20020314	rain_SI_20020314	rainSE_SI_20020314	IKind20020314	IKrain20020314
20020312 00-24 UTC	rain_20020313	rain_SI_20020313	rainSE_SI_20020313	IKind20020313	IKrain20020313
20020311 00-24 UTC	rain_20020312	rain_SI_20020312	rainSE_SI_20020312	IKind20020312	IKrain20020312
20020310 00-24 UTC	rain_20020311	rain_SI_20020311	rainSE_SI_20020311	IKind20020311	IKrain20020311
20020309 00-24 UTC	rain_20020310	rain_SI_20020310	rainSE_SI_20020310	IKind20020310	IKrain20020310
20020308 00-24 UTC	rain_20020309	rain_SI_20020309	rainSE_SI_20020309	IKind20020309	IKrain20020309
20020307 00-24 UTC	rain_20020308	rain_SI_20020308	rainSE_SI_20020308	IKind20020308	IKrain20020308

Work in progress includes the incorporation of gauge-tuned radar and satellite rainfall estimates. These are expected to provide improved accuracy and detail in regions with few gauges.

For more information on analyses produced using Statistical Interpolation:

Weymouth, G.T., M.M. Manton, and D.A. Jones, 1999: Analysis of rainfall from gauge observations - accuracy and statistical interpolation. BMRC Research Report, manuscript in preparation.

For more information on the GMS "no rain" augmented analyses:

Ebert, E.E. and G.T. Weymouth, 1999: Incorporating satellite observations of 'No Rain' in an Australian daily rainfall analysis. Journal of Applied Meteorology, 38, 44-56.

These results are experimental and are not part of the Bureau of Meteorology's operational products and services.