Please enable JavaScript.
Coggle requires JavaScript to display documents.
Five weather forecasting challenges (How much MetOffice is worth (Land…
Five weather forecasting challenges
How much MetOffice is worth
Land transport: £100m per annum
Avoiding flood damage: £64m per annum
Public sector & aviation: both have value of £400m per annum
Added value to economy: £400m
Sum benefits around £1.5bn per annum
Numerical models
Predict
way atmosphere is going to evolve- these 3 equations are used together to
describe the flow of a fluid in the atmosphere
:
Take Navier- Stokes equation (conservation of momentum) on a rotating sphere
Conservation of energy: change in temp of the system is related to its sources and sinks
Continuity equation: represents the conservation of mass
How it's done
Split the sphere into a number of different rectangualr boxes w/ their own conditions (wind, temp, pressure, moisture)
For each grid
work out rates of change
for each factor (humidity, temp) by solving the equations of motion numerically
Take the new atm state at the new time and solve for the evolution over the next time-step. Repeat
A very large computer is needed: E.G: Week's forecast using resolution of 17km (100 million boxes) in under an hour
Timestep typically around 10 mins for global model
Clouds
Droplets
Well defined mass
Well defined fall speeds
To represent clouds in model: blow air over (in from left)-
pressure decreases
over the mountain and cools to reach
water saturation
where cloud forms
In a front: at surface we experience rain, however rain is caused from ice above, which comes to freezing level, melts -> rain
Mass, terminal fall speed is known for rain but less known for ice (diff shapes)
In Britain: 1/2 to 2/3 of the time, the rain experienced at the surface is due to snow melting
Ice
Multitude of different shapes
No well defined mass
No well defined fall speeds
In a model: need to predict way water moves between different phases, shown by
cloud microphysical scheme
Rain -> Vapour -> Liquid -> Ice
Starting conditions
In order to make forecast, the starting conditions all around the world
at one time
is required
Lots of
surface observations
data over US and Europe, sparse data over Africa, oceans and parts of Antarctica
Satellite observations
have a more even spread across the world
Data assimilation:
dropsondes, ozone SCATT radiances, GPS satellites, TEMP, SYNOP-ship, aircaft
Data assimilation
Data assimilation in weather forecasting
does NOT provide an exact initial state of the atm
based on observations
Data assimilation provides a
best-guest initial state
of atm consistent with the model
Provides
estimates of model-obs errors
Chaos & uncertainty
Small changes in initial conditions can lead to large differences later
Use of global ensemble system
Use all possible solutions as starting points to the model- used to figure out if something is certain to happen/ less certain
Communicating the forecast
Use of likelihood & impact table
Orange: med to high
Red: very likely & high impact- hurricane possible
Yellow: medium to low- be aware
Green: no/ low impact
Probbility low impact
probability medium impact
Probability high impact
Wind warnings can be used with rain warnings