F6, LOL, wish i'd caught that whenever shortly after 1946. Could have changed my whole life. Maybe order from random or some order from mostly random or even more order from almost total random, so many options. If only i'd caught it back then.
This caught my eye in the word section of your post ..
These applications, called Monte Carlo methods, required a large supply of random digits and normal deviates of high quality, though i did question the "normal" .. WHY? did it catch my thought? Will leave that to your imagination, to preserve this post. lol
Seriously yours did remind me of one in my weekend paper which i tore out for you and which
the need arose for random numbers that could be used to solve problems of various kinds of experimental probability procedures.
reminded me of .. (mine was in another paper)
Lightning strikes: you do the maths DAMIEN MURPHY August 30, 2009
Lightning over Mt Dandenong. Photo: Robert Carew
Ever thought it would be handy to know where a severe thunderstorm would hit? Now you can, and it's a simple as …
Well, not that simple, really.
The Thunderstorm Environment Strike Probability Algorithm, or THESPA, is a radar-based system which uses mathematical data to estimate strike probability in a defined area as a storm starts to build or move.
Developed by Sandy Dance at the Centre for Australian Weather and Climate Research, it can produce 'nowcasts' for up to one hour ahead.
'Based on probability, we can be far more confident in estimating the … geographic point where the thunderstorm is likely to strike,' Dr Dance said. 'It will prove invaluable in helping forecasters with risk assessment and emergency decision-making.'
The system was developed using records of thunderstorms near Sydney which were then applied to data collected from 12,597 storms in China tracked over a 51-day period around last year's Beijing Olympics.
Existing methods permit stormchasers to predict only a 'threat area' where a storm is likely to appear and their predictions were 'pretty much ad hoc, flying by the seat of the pants stuff'.
The algorithm system uses radar to bounce off water droplets in a thunderstorm and compute the probability a given location will be affected by a storm.
'Previously the nowcast was of a deterministic nature, often with an implicit threat area indicated by velocity vectors and detected storm shapes,' Dr Dance said.
'Forecasters must then interpret these indicators to make public warnings, acknowledging that the nowcast usually contains significant errors in position, size and intensity, owing to uncertainties in storm motion and evolution.'
The Bureau of Meteorology defines a severe thunderstorm as one that produces hail the size of a $2 coin, gusts of wind exceeding 90km/h, flash floods and tornadoes. Thunderstorms develop when warm, humid air near the ground receives an initial upward push from converging surface winds and rises rapidly in an unstable atmosphere. Severe thunderstorms are localised and affect areas much smaller than tropical cyclones or floods.
Based on insurance payouts, however, the bureau said severe thunderstorms cause more damage each year than tropical cyclones, earthquakes, floods or bushfires.
They can also be lethal. Lightning strikes kill between five and 10 Australians a year and more deaths occur when strong winds cause tree limbs to fall, turn debris into projectiles or capsize boats in open water.
Predicting the path of a storm (Image: Top Photo Group/Rex Features)
Chasing thunderstorms just got easier. A new system for predicting where they will go next will soon be in use in Australia, after outperforming existing methods. It will help forecasters decide whether to give public warnings of impending storms.
The technique, called Thunderstorm Environment Strike Probability Algorithm (THESPA), produces "nowcasts" for up to 1 hour ahead. It was developed by a team led by Sandy Dance at the Centre for Australian Weather and Climate Research. ................................................. INSERT .. .. aha! .. finally found the algorithm .. Sandy Dance .. http://www.cawcr.gov.au/staff/sandy/index.php .. 2nd down in Publications, hope it was .. bottom page 4 .. http://www.cawcr.gov.au/staff/sandy/iemss/thespa_iemss.pdf ................................................. THESPA borrows an idea used to predict the movements of cyclones – much bigger weather systems that include hurricanes and typhoons. It assigns a strike probability – the chance that a particular location will be hit by a storm in a given period.
"The concept of a strike probability is new for thunderstorms," says Dance.
Storms on the move
Existing forecasting computer systems use broad-brush methods: for instance, trying to predict overall stormy conditions over a region. They show a "threat area" that thunderstorms are likely to pass through.
Like all weather forecasts they are not completely reliable, but crucially they do not tell forecasters how far wrong the prediction is likely to be. So the forecasters must take the predicted threat area and draw in, by hand, a slightly wider region. They then issue a thunderstorm warning for this larger area.
THESPA's use of strike probabilities removes this element of guesswork: it detects individual storms and predicts their paths. As a result, Dance says, "Our system is more geographically specific." Thundering good
The system was "trained" using records of thunderstorms near Sydney, Australia. It was then tested on a database of 2394 storm tracks, which were recorded in the Beijing area of China around the time of the 2008 Olympics.
The system performed well, achieving a Brier Skill Score of 0.39. On this scale, 1 would be a perfect score, and 0 would be as good as the present manual forecasting method.
Clive Wilson of the UK Met Office says the system looks promising. "They seem to be getting better results than the existing methods, and it's obviously practicable," he says. "But they need to run it on a bigger sample to check that it really does score this well."
The Australian Bureau of Meteorology will start using THESPA for forecasts in the near future.
Journal reference: Journal of Atmospheric and Oceanic Technology, DOI: 10.1175/2009jtecha1279.1 http://dx.doi.org/10.1175/2009JTECHA1279.1 www.newscientist.com/data/images/ns/cms/dn17681/dn17681-1_300.jpg
News 
Market Data 
Discover 
