BotB

ActuallyLTN Phobia, you raise (no pun intended ) a very good point...

People ask 'why is it so hard to forecast the weather?'...


20.05.12 BA0233 East of Moscow from Prospero

This picture illustrates the problem perfectly....why in a fairly small patch of Cumulus cloud is there all of a sudden this fantastic and aggressive growth occurring? What perfect conditions exist in that one place at this moment in time to allow such development of the cloud here but no where else?...The answer is we don't know exactly but are trying to find out...So how do we do this...

We find special cases like this particular situation and then try to get as much data as possible about what the conditions were like...This equates to collecting (where possible) up to 100 variables such as temperature at the surface, at 10m, at 100m, etc..., humidity at the same levels, pressure, wind, cloud cover, solar heating at several bandwidths (short, long, total, direct, indirect), aerosol particle counts, gas constituents, etc...then we look at the geography at this spot and try to then put the specific lifting, heating, cooling convergence, divergence, etc....actions into the computer given this landscape and then run one simulation against all other variations (adjusting the weighing of each parameter against the others) to find what scenario creates this development that nature has made...then we try to repeat this on other similar areas of landscape and heating, season, moisture, etc...to see if they do repeat predictably...if they do we create what is known as a 'local model' that we apply to the super computer when it is running the various scenarios for the weather in the UK...taking into account (where possible and known) all of the 100 variables and all of the local geographic variations and the effects these variations will have on the weather to become modified....You can imagine the amount of 'runs' of the simulations that are needed to figure out which scenarios are most likely and seem to repeat (giving a greater probability)...

So we are effectively trying to understand the very small scale, small scale, mid size scale and large scale effects and how they all interact together with the various 100 or so 'modifiers' such as temperature at a specific height, solar radiation at a specific location at a specific season or day...to see what the best 'gamble' is...due to percentages as we don't yet know all of the interactions and outcomes for each specific scenario...but we are learning them more and more quickly and able to forecast them better with more and more computing power...

To give you an idea...we are now able to run 'local scale' grids for the computer models down to roughly 1.5 Km cubed...Given the diameter of the Earth that is staggering really...of course we don't run the whole Earth model at this fine scale all at once...but we do for the UK for example and that as often as possible to keep up with the current conditions that we have measured and try to ensure we are weighing the correct parameters with the correct scores...

It all boggles the mind really...

So, very isolated, extreme events such as this one TCU cloud in an otherwise very normal CU cloud deck does make for some interesting science ^

However, you are correct...it would indeed be quite 'rude' to fly through that cloud at that particular time as the updraft would be quite decent I think...again the popcorn analogy works well here...just on a bigger scale...

Sorry, gone off on another one! Right off to dreamland then...