The BOM’s weather modelling is abysmal. It is nearly as bad as their climate modelling.
Are you sure this isn’t the BoM’s April fools day joke Warwick?
Sure did not get the prediction right for SE Qld My place is in the 35-40% of median. I do not think they got the actual chart right either. My measurement was 481mm (185%) compared with the long term average of 260mm. Other place in SE Qld had higher rainfall and there is prediction of more flooding. In some areas there has been more flooding than 1974 and 2011. It is lucky that the water people got it right this time and had dam levels lower with controlled releases to accommodate the flood surge so less flooding in Brisbane, Gympie and Maryborough
Warwick, how do we compare rainfall as a percentage of the mean to the predicted chance of rain? I don’t know how to judge the success of the prediction. The colours change a lot between the two graphs but I don’t know what it means. The colour legends are quite different, and what is depicted in each seems incompatible. If this is normal then I’m sorry, but I’m still confused. Thanks.
In the case of March that I have commented on – obviously the BoM prediction had a good win in that large brown area in Central Oz plus the Western Pilbara on the Percent Rain map.
Other areas which experienced rain over the mean were big fails.
The Outlook predicted wetter conditions on a tiny area on tip of Cape York and even that was wrong as it turned out only 60 – 80% of the mean.
It would be easy to devise a scoring system for the BoM gridded data but it would require an equal number of divisions in the scales.
‘how do we compare rainfall as a percentage of the mean to the predicted chance of rain? I don’t know how to judge the success of the prediction. ‘
The BoM outlook is the chance of exceeding the median rainfall (note not average rainfall). This is the least precise rainfall prediction possible. Ignoring the fact, the percentages are spurious. They are just artifacts of how the models are run. All the outlook says is some places will have more rain than the median. Other places less. And will be accurate for half of Australia by chance.
Bear in mind that precision and accuracy vary inversely. The less precise the forecast, the more accurate it will be, and the chance level for the BoM outlook being accurate is 50% of the country. So the accuracy is rather less impressive than it might appear.
Phillip B, I suspect that at times you are an apologist for BOM.
I suggest that BOM has a very poor record for the following reasons
1/ They do not look sufficiently at past records particularly those long term sites that were established in the 19th century eg Gayndah Qld.
2/ They do not look for patterns in relation to things like sun spots, sun spot cycles (approx 11 years) position of planets (Jupiter’s orbit is about 11 years), orbit of earth around the sun, position of the moon (tides, including king tides are calculated from the orbits of the moon)
3/ They do not recognise UHI and correct records the wrong way – ie they reduce past temperatures where there was no UHI and adjust upwards temperatures where there is now UHI (instead of reducing the temperatures by the extent of UHI which can be as much as 6C)
4/ They have the wrong models which includes allowance for CO2 when as any competent chemical or mechanical engineer knows CO2 (from measured experimental data) has no influence on atmospheric temperature
5/ The wrong harmonise records from sites that have no relation to the climate & geography of other sites.
6/ BOM has abandoned many good rural sites for airport sites which have a poor record of reliable data (eg UHI, blasts from jets, local turbulence etc)
7/ weather generally moves from west to east across the continent but BOM with satellites and radar still can not get short term predictions correct predictions. Sometimes just looking out of the window gives a lie to the BOM weather forcast on the TV, radio or on the internet.
8/ the people at the top of BOM who have political leanings and dictate policy should be sacked.
I’m surprised you consider me an apologist for the BoM.
For the record, I think surface warming is mostly a spurious signal from using minimum temperature to determine an average temperature. If the claimed warming were real, we would see it in the Australia measurements taken at fixed times, which go back to the 1950s, and we don’t.
It’s as simple as reduced lowlevel clouds from decreased aerosols causing earlier and higher minimum temperatures. Which is not to say UHI isn’t real, but it is in large part an aerosol/cloud phenomena as well.
I thought the recent study that showed less than 1% of papers followed the scientific method timely to say the least. Incidentally, someone I know well and is a very prominent scientist told me essentially the same thing 20 years ago.
Leave a Reply
Primarily exposing faulty methodologies behind global temperature trend compilations
The BOM’s weather modelling is abysmal. It is nearly as bad as their climate modelling.
Are you sure this isn’t the BoM’s April fools day joke Warwick?
Sure did not get the prediction right for SE Qld My place is in the 35-40% of median. I do not think they got the actual chart right either. My measurement was 481mm (185%) compared with the long term average of 260mm. Other place in SE Qld had higher rainfall and there is prediction of more flooding. In some areas there has been more flooding than 1974 and 2011. It is lucky that the water people got it right this time and had dam levels lower with controlled releases to accommodate the flood surge so less flooding in Brisbane, Gympie and Maryborough
Warwick, how do we compare rainfall as a percentage of the mean to the predicted chance of rain? I don’t know how to judge the success of the prediction. The colours change a lot between the two graphs but I don’t know what it means. The colour legends are quite different, and what is depicted in each seems incompatible. If this is normal then I’m sorry, but I’m still confused. Thanks.
Gidday Richard, You can map Australian rain by various parameters and timescales here –
www.bom.gov.au/jsp/awap/rain/index.jsp
The Outlooks can be obtained at their map archive here –
www.bom.gov.au/jsp/sco/archive/index.jsp?map=rain
In the case of March that I have commented on – obviously the BoM prediction had a good win in that large brown area in Central Oz plus the Western Pilbara on the Percent Rain map.
Other areas which experienced rain over the mean were big fails.
The Outlook predicted wetter conditions on a tiny area on tip of Cape York and even that was wrong as it turned out only 60 – 80% of the mean.
It would be easy to devise a scoring system for the BoM gridded data but it would require an equal number of divisions in the scales.
‘how do we compare rainfall as a percentage of the mean to the predicted chance of rain? I don’t know how to judge the success of the prediction. ‘
The BoM outlook is the chance of exceeding the median rainfall (note not average rainfall). This is the least precise rainfall prediction possible. Ignoring the fact, the percentages are spurious. They are just artifacts of how the models are run. All the outlook says is some places will have more rain than the median. Other places less. And will be accurate for half of Australia by chance.
The Bom do publish maps of outlook accuracy,
www.bom.gov.au/climate/ahead/verif/
Bear in mind that precision and accuracy vary inversely. The less precise the forecast, the more accurate it will be, and the chance level for the BoM outlook being accurate is 50% of the country. So the accuracy is rather less impressive than it might appear.
Phillip B, I suspect that at times you are an apologist for BOM.
I suggest that BOM has a very poor record for the following reasons
1/ They do not look sufficiently at past records particularly those long term sites that were established in the 19th century eg Gayndah Qld.
2/ They do not look for patterns in relation to things like sun spots, sun spot cycles (approx 11 years) position of planets (Jupiter’s orbit is about 11 years), orbit of earth around the sun, position of the moon (tides, including king tides are calculated from the orbits of the moon)
3/ They do not recognise UHI and correct records the wrong way – ie they reduce past temperatures where there was no UHI and adjust upwards temperatures where there is now UHI (instead of reducing the temperatures by the extent of UHI which can be as much as 6C)
4/ They have the wrong models which includes allowance for CO2 when as any competent chemical or mechanical engineer knows CO2 (from measured experimental data) has no influence on atmospheric temperature
5/ The wrong harmonise records from sites that have no relation to the climate & geography of other sites.
6/ BOM has abandoned many good rural sites for airport sites which have a poor record of reliable data (eg UHI, blasts from jets, local turbulence etc)
7/ weather generally moves from west to east across the continent but BOM with satellites and radar still can not get short term predictions correct predictions. Sometimes just looking out of the window gives a lie to the BOM weather forcast on the TV, radio or on the internet.
8/ the people at the top of BOM who have political leanings and dictate policy should be sacked.
I’m surprised you consider me an apologist for the BoM.
For the record, I think surface warming is mostly a spurious signal from using minimum temperature to determine an average temperature. If the claimed warming were real, we would see it in the Australia measurements taken at fixed times, which go back to the 1950s, and we don’t.
It’s as simple as reduced lowlevel clouds from decreased aerosols causing earlier and higher minimum temperatures. Which is not to say UHI isn’t real, but it is in large part an aerosol/cloud phenomena as well.
I thought the recent study that showed less than 1% of papers followed the scientific method timely to say the least. Incidentally, someone I know well and is a very prominent scientist told me essentially the same thing 20 years ago.