Experimental Forecasts
Below are some very good but experimental forecast links. If you just want to quickly check them out now, feel free. There is also a BoM Beta forecast link in the menu above. If you want to learn how to get more value out of these experimental forecast links and become an amateur forecaster yourself, read the blurb that goes with each link. All links here are experimental in nature without official forecaster input so please keep that in mind when using them. However, IMO, they can be notably better than an app. The ensemble links, in particular (lower down the page) are very good. Use the percentiles (ie: 90% and 10%) to grasp rainfall potential.
Rain Forecast + Rapid ECMWF hourly forecast. - This Meteologix website is brilliant. Run your mouse (or your finger on a phone) from left to right on the rainfall graph. As you move from left to right, you will get a rain forecast (a collection of various models, which are the coloured dots) with each model giving you a total in mm for the time indicated. You can de-select those flag icons below, too. They are the model colour keys.
I'd try only using the top three flags (EC, EC and USA) and the aussie flag - as these models are the best. De-select the others by pressing them. USA and Aussie are best for storm prediction. EC is best for general forecasting. At the very end of the graph, you will get an accumulation forecast from each model for either two to three days or up to ten days - depending on which tab you select at the top.
Scroll down below the rainfall graph and you will see the option to select a particular model (ECMWF Rapid and ECMWF HD and GFS/USA, are typically the best) and then view the weather from hour to hour below that selection. Scroll to the very bottom of the page, to get a model time stamp, which will tell you what model time release you have. Add 10 hours (11 in daylight savings) to the model time stamp, to see what time that model was released in AEST time. Experiment with the models and see what results they give. Remember, ECMWF (ECMWF Rapid updates four times a day) is typically the best, along with GFS, although GFS can over estimate rainfall but can be better at predicting instability and storms across the first 24 hours of a forecasting period. This link is based on model forecasts with no human input, so whilst it can be more accurate (IMO) at times than the official BoM forecast, you need to be aware that it is not an official forecast. NB: Temperature forecasts are not as accurate at times as other aspects of this initiative.
Meteologix Katoomba Meteogram - This is another excellent presentation from Meteologix. This meteogram link gives you many graphs for your finger or mouse to view from left to right, as the wind, sunshine, precipitation graphs etc reveal their forecasts. The cloud cover graph is just sensational, showing you forecasts for low, mid and upper level cloud on a given day. Select different models for different estimates of the weather variables on this link and for the best forecasts, stick with ECMWF Rapid (only goes out to +72 hours) ECMWF and GFS (USA) models for best results.
ECMWF Rapid is released four times a day (roughly 6am, midday, 6pm and midnight) and with its hourly data input, has some quite detailed and high resolution information although only out to 72 hours. The normal ECMWF is excellent for short and much longer term forecasts and is often used by the BoM. It does sometimes underestimate rainfall - IMO - especially during showery/stormy weather whereas the GFS (USA) model can over-estimate rainfall but is good at suggesting storm potential in the first 24 hours of its forecast. This is where human forecasters can come in, to add some intuition and experience to these model forecasts. NB: Temperature forecasts are not as accurate at times as other aspects of this initiative.
Meteologix Ensemble Forecast Katoomba - 'Ensemble forecasting is a method used in numerical weather prediction. Instead of making a single (control) model forecast of the most likely weather, a set (or ensemble) of up to fifty slightly different model (perturbations) forecasts are produced. This set of forecasts aims to give an indication of the range of possible future states of the atmosphere.' - Wikipedia. So, compare the ensemble forecast with the deterministic outlook. It takes some knowledge and experience to use this feature but some of you might find it interesting. Use the 90% percentile and 10% percentile to get an idea of the higher and lower end of the forecast range and the 'main run' to see what the determistic (singular/control) model outlook is.
Ensemble forecasting - Explained in simple english.
Meteologix Ensemble Forecasts - Search Box - 'Ensemble forecasting is a method used in numerical weather prediction. Instead of making a single forecast of the most likely weather, a set (or ensemble) of forecasts is produced. Use this search box for your town or suburb and then slide your finger/mouse across the graph to get a percentile output. Use the tabs for temperatures, wind etc.