| |
Numerical weather prediction |
Numerical weather prediction concentrates upon two problems:
diagnosing the current state of the atmosphere and numerically
modelling how the atmosphere will evolve with time.
Many different research groups in the Met Office work
together with the overall aim to continually improve the operational NWP forecasts; in
turn increasing the accuracy of the forecasts that are issued. The
model may be improved by better numerical techniques and representation
of the atmosphere, by making better use of existing observations or by
developing new observation types.
INITIAL CONDITIONS
| OBSERVATIONS |
Observations are crucial to weather forecasting. Many thousand
are received each day and these are processed, quality controlled
and monitored. Within this area you may learn more about
different observation types, the daily coverage and the
processing that is performed before they are used. Another aspect
is re-running forecasts to determine what effect particular
observations had.
More about how
observations are used
|
| DATA ASSIMILATION |
After observations have been processed and quality controlled
they then need to be incorporated, or assimilated, into the
numerical model. In doing this very important process, we get a
representation of the current state of the atmosphere, i.e. an
analysis, from which a forecast is obtained.
More about data analysis
and assimilation
|
| SATELLITE APPLICATIONS |
Perhaps one of the greatest developments in weather forecasting
over the past 40 years has been the development of weather
satellites. A satellite image of current cloud conditions is a
familiar sight on TV weather forecasts but there is far more to
satellite data than simply producing photographs. Here you will
learn about the many ways that satellite data are used from an
ever-growing range of instrumentation.
More about satellite
applications
|
| ENSEMBLE FORECASTS |
Typically, a numerical weather prediction model is only run once
from a given set of initial conditions to produce a single
forecast. However, despite the vast improvements in NWP models
over the years, large errors may still occur even over relatively
short forecast ranges. This is due to the chaotic nature of the
atmosphere which means that a small error in the initial
conditions may lead to a large error in the subsequent forecast,
the so called 'butterfly effect'. Because we can never know the
exact truth of the initial conditions, there will always be a
degree of uncertainty in the resulting forecast. To combat this,
an ensemble suite of forecasts may be run with each forecast
having slightly different initial conditions to reflect the
uncertainty. The resulting forecasts may be studied and the
possible range of different scenarios evaluated. This may be used
to calculate a probability for a particular forecast sequence.
More about ensemble
forecasts
|
NUMERICAL MODELLING
|
Atmospheric numerical modelling is the process of solving a set
of equations to obtain an objective forecast of the future state
of the atmosphere. The equations describe the evolution of many
variables (e.g. temperature, wind speed, humidity and pressure)
and together define the state of the atmosphere. Numerical models
may be very complicated, consisting of numerical integration
schemes, physical parametrizations and data output schemes.
More about numerical
modelling
Forecasting with high-resolution models
|
| NUMERICS |
All numerical models of the atmosphere are based upon the same
set of governing equations. The numerical models differ in the
approximations and assumptions made in the application of these
equations.
More
about the equations used in the Met Office models
|
| ATMOSPHERIC PROCESSES AND
PARAMETRIZATIONS |
Many of the physical processes that occur within the atmosphere
are at a scale too small to be directly resolved by weather
forecasting or climate models. Therefore, numerical
representations of significant effects of these processes must be
developed: we call these representations 'parametrizations'. In
order to parametrize the processes we need to understand them. We
do this by undertaking research using observations and
very-high-resolution numerical models.
More about atmospheric
processes and parametrizations
|
| OBSERVATIONAL BASED
RESEACRH |
The observational based research is based on data from two
facilities; an aircraft and a large range of surface-based
instrumentation including a tethered kite balloon. Both the
aircraft and the balloon are heavily equipped with
instrumentation for determining standard meteorological
parameters and more-specialised instrumentation for measuring
turbulence, cloud, aerosol, radiation and chemistry.
More about
observational based research
|
|
|
|
Home 
