Assignment 1

Question 1

What is Numerical Weather Prediction? Give a brief explanation of the meaning of numerical models.

Answer 1

1. Numerical Weather Prediction: Part of meteorology science, which is dealing with modeling of the atmospheric conditions and trying to predict these conditions in the near future. (24 hours up to 7 days)
   
   1. Among the most expensive technologies in term of CPU consumption.​
2. Numerical models: equations installed and processed using computers to predict and forecast the weather.

Question 2

State the 3 models used in the GHQ Met department

Answer 2

1. WRF (Weather Research and Forecasting) model for weather parameter prediction.
2. WAM (Wave Model) for sea state forecast.
3. WRF-CHEM for pollutants and dust transport.

Question 3

Differ between hydrostatic and non-hydrostatic model

Answer 3

1. Hydrostatic approach assumes hydrostatic equilibrium, in which the downward weight of the atmosphere balances the upward-directed pressure gradient. (Laminar movement of air masses in the middle of the atmosphere) no connection between vertical and horizontal movement.
2. Non-hydrostatic processes and their effect become important when the length of a feature is approximately equal to its height. (Mesoscale) usually happen at the surface.
   1. Convective storms. (Mesoscale)
   2. Gust fronts.
   3. Gravity waves (mountain waves and turbulence)

Question 4

Why there is a necessity to parameterize some atmospheric phenomena?

Answer 4

1. Computers are not yet powerful enough to directly treat them.
2. They are often not understood well enough to be represented by an equation.
3. Their effects profoundly impact model fields and are crucial to creating realistic forecasts

Question 5

What is the process of Data Assimilation?

Answer 5

It is the process of comparing an old forecast with the new observation where the validity of the forecast is checked. The model state then is updated and a new forecast is initiated. (sequential time-stepping procedure)

Question 6

What is the difference between Conventional and Non Conventional observations?

Answer 6

1. Conventional observations: observations taken from stations or taken by instruments those are actual data recorded at real time and transmitted through the global telecommunication system such as SYNOP, SHIP, BUOY, SYNOR, TEMP, PILOT, AIREP, AMDAR, ACAR, SATEM, SATOB.
2. Non conventional observations: observations/ imagery that are typically taken by atmospheric remote sensors “no physical contact between the sensor and earth’s surface or phenomena of interest” or they are not actual data they are usually estimated from GTS data therefore used with the support of conventional data. They are not accurate and the possibility of using non conventional data alone may result in errors in the prediction model. examples of non conventional data: RADAR, TOVS, SSMI, …etc.

Question 7

What is the difference between a LAM model and a Global model?

Answer 7

1. LAM (limited area models):
   
   1. Resolve atmospheric equations on regional or local limited area domains.
   2. They can use very high resolution (100m to 50 km) and more vertical levels.
   3. Can catch very small phenomena.
   4. Can run on small to medium computers (normal PCs, workstations, clusters, servers).
   5. Obligated to get LBC and initial data from global models.
2. Global models
   
   1. Resolve atmospheric equations on the whole globe.
   2. Cannot use very fine resolution because of computer limitations.
   3. Cannot detect small-scale phenomena.
   4. Need powerful computers (CRAY, FUJITSU VPP, NEC SX, IBM, SGI, …etc).
   5. Used to forecast general synoptic circulation and provide initial and lateral boundary data for limited area models

Question 8

Give a definition of the LBCs (Lateral Boundary Conditions) ? Why are they necessary for a LAM model?

Answer 8

1. Definition
   
   1. LBC Come from global or regional models including the LAM domain
   2. Data downloaded from the Internet; they are generally very heavy (large size)
      
      1. GFS ~ 3.5 GB per day (0.5 degree horizontal resolution, 25 vertical levels)
      2. WAFS ~ 120 MB per day (1.25 degrees resolution, 11 vertical levels).
2. Why are they necessary for a LAM model?
   
   1. To be able to compute derivatives at the model borders.
   2. To know what is likely to penetrate the domain of interest from outside.
   3. To avoid noisy fields at the borders.

Question 9

Explain the meaning of both Model Dynamics and Model Physics?

Answer 9

1. Model dynamics
   
   1. Represent large scale motions of the atmosphere in response to large scale gradient of temperature and pressure together with earth’s rotation for example (conservation of momentum, mass, energy and water)
   2. Concerned about the equations only
2. Model physics
   
   1. All other processes that occur in the atmosphere on smaller scales but which are nonetheless essential to the evaluation of weather
   2. Pertain to source and sinks of the same quantities which the dynamical equations are describing the conservation
   3. Part of model correct the dynamic terms (part of equations)

Question 10

Why do we need HA and HPC resources to run numerical models?

Answer 10

1. High availability
   
   1. Insure that processes and tasks are being done all the time if one stops another program continues the job (redundancy)
2. High performance computing
   
   1. Can take a big load of work therefore need a large amount of processors to solve equations. This amount of processors can only work on a supper computer.

Question 11

Give some examples of known international Global and LAM models?

Answer 11

1. Global models
   
   1. ECMWF/IFS
   2. NCEP/GFS
   3. Metro-France/ARPEGE
   4. German DWD global model
   5. METOFFICE/UKMO Unified Model
   6. Japan Meteorological Agency Global Model
2. Limited Area models
   
   1. WRF/NMM (Weather research and Forecasting)
   2. ALADIN
   3. COSMO/LM
   4. HIRLAM
   5. COAMPS, RAMS, RUC, … etc.
   6. ETA
   7. MM5