Numerical Weather Prediction
CSIR NET Part C meteorology: computational instability, filtering, NWP models, forecast ranges, objective analysis, data assimilation, satellites.
- 4
- Topics
- 85
- Min read
- 01 Computational Instability and FilteringStart here CSIR NET meteorology: CFL condition, von Neumann analysis, time schemes, leapfrog computational mode, Robert–Asselin filter, aliasing and filtered equations.
- 02 Barotropic, Baroclinic and Primitive-Equation Models CSIR NET meteorology: barotropic model, Rossby–Haurwitz waves, quasi-geostrophic baroclinic models, primitive equations, σ coordinates, spectral models.
- 03 Short-, Medium- and Long-Range Prediction CSIR NET meteorology: forecast ranges, predictability and chaos, error growth, ensemble prediction, seasonal monsoon forecasts and verification scores.
- 04 Objective Analysis, Initialization, Data Assimilation and Satellite Data CSIR NET meteorology: Cressman and optimal interpolation, normal-mode and digital-filter initialization, 3D-Var, 4D-Var, Kalman filters and satellite data.
- Numerical Weather Prediction Short Notes Revise The whole chapter on one page.
Chapter Overview
Numerical Weather Prediction (NWP) is a unit of the Meteorology section of Part C in CSIR NET Earth, Atmospheric, Ocean and Planetary Sciences (subject code 702). It asks how a forecast is actually made: how the equations of atmospheric motion are turned into a computer program that stays stable, which simplified or full equation sets are integrated, how far ahead each kind of forecast can be trusted, and how scattered observations become the initial state the model starts from.
The unit sits between dynamics and practice. It uses the vorticity equation, quasi-geostrophic theory and the hydrostatic primitive equations, but its questions are about numbers and choices: a time step, an amplification factor, a Rossby wave speed, an error-doubling time, a verification score, an optimal-interpolation weight. Every numerical answer on these pages has been recomputed and is worked in full.
The notes assume an MSc-level course in atmospheric dynamics and numerical methods. Standard references for the material are Haltiner and Williams’s Numerical Prediction and Dynamic Meteorology, Kalnay’s Atmospheric Modeling, Data Assimilation and Predictability, and Holton’s An Introduction to Dynamic Meteorology.
What You’ll Learn
By the end of this chapter you will be able to:
- Test a scheme with von Neumann analysis, apply the CFL condition, and compare forward, backward, trapezoidal, Matsuno and leapfrog time differencing
- Explain the leapfrog computational mode and the Robert–Asselin filter, aliasing and nonlinear instability, and which approximations filter sound and gravity waves
- Contrast barotropic, quasi-geostrophic baroclinic and primitive-equation models, compute Rossby–Haurwitz wave speeds, and use σ coordinates and spectral truncation
- Place a forecast in the WMO ranges, reason about chaos and error growth, read an ensemble, and score forecasts with RMSE, ACC, POD, FAR, CSI and the Brier score
- Compute Cressman and optimal-interpolation analyses, compare initialization methods, 3D-Var, 4D-Var and EnKF, and explain how satellite radiances enter NWP
Prerequisites
- Finite differences, stability and least squares: Numerical Analysis and Inversion (Geophysics), especially Numerical Differentiation, Integration, FDM and FEM and Well- and Ill-Posed Problems and Least Squares
- Hydrostatics, thermodynamics and radiative transfer: Physical Meteorology
- Part B background on the general circulation, monsoon and ENSO: Oceans and Atmosphere
- The mathematicians’ numerical methods: Numerical Analysis (Mathematical Sciences)
Learning Path
The topics follow the order of the official syllabus lines, which is also the logical order: numerics, then models, then what models can forecast, then how they are started.
graph TD
A["1. Computational Instability
and Filtering
CFL, time schemes, aliasing,
filtered equations"] --> B["2. Barotropic, Baroclinic and
Primitive-Equation Models
vorticity, QG, PE, σ, spectral"]
B --> C["3. Short-, Medium- and
Long-Range Prediction
chaos, ensembles, seasonal,
verification"]
A --> D["4. Objective Analysis,
Initialization, Data Assimilation,
Satellite Data"]
B --> D
D --> C
C --> E["Quick Revision"]
D --> EPage 1 explains why models need short time steps and filters. Page 2 shows which equations are integrated and what each version can forecast. Page 4 builds the initial state, and its initialization section completes the story of gravity-wave noise begun on page 1. Page 3 ties everything to forecast skill at each lead time; read it last if you prefer.
Topics in This Chapter
| # | Topic | What it covers | Time |
|---|---|---|---|
| 1 | Computational Instability and Filtering | consistency, stability, Lax theorem, von Neumann analysis, oscillation and advection equations, forward, backward, trapezoidal, Matsuno, Heun and leapfrog schemes, computational mode, Robert–Asselin filter, CFL condition, upstream scheme, semi-implicit and semi-Lagrangian methods, pole problem, phase errors, staggered grids, aliasing and nonlinear instability, Arakawa Jacobian, filtered equations | 40 min |
| 2 | Barotropic, Baroclinic and Primitive-Equation Models | barotropic and baroclinic atmospheres, barotropic vorticity equation, ENIAC forecasts, Rossby–Haurwitz waves, equivalent barotropic and shallow-water models, QG equations and potential vorticity, two-level model, primitive equations, prognostic and diagnostic variables, vertical coordinates, grid-point versus spectral models, parameterizations, limited-area models | 40 min |
| 3 | Short-, Medium- and Long-Range Prediction | WMO forecast ranges, chaos, intrinsic and practical predictability, error doubling, predictability of the first and second kind, short-range tools and MOS, ensemble prediction, extended range and MJO, seasonal monsoon forecasting, verification scores | 35 min |
| 4 | Objective Analysis, Initialization, Data Assimilation and Satellite Data | background and observation errors, polynomial fitting, Cressman and Barnes, optimal interpolation and BLUE, static, dynamic, normal-mode and digital-filter initialization, 3D-Var, 4D-Var, Kalman filter and EnKF, quality control, polar and geostationary satellites, radiances, weighting functions, retrievals, AMVs, scatterometers, radio occultation | 40 min |
| — | Quick Revision | every formula, range and trap on one page | 10 min |
Official Syllabus Lines and Where They Are Covered
| Official topic line | Covered in |
|---|---|
| Computational instability, filtering | Computational Instability and Filtering (numerical filters and filtered equations); gravity-wave filtering by initialization in page 4 |
| Barotropic and baroclinic models, primitive-equation models | Barotropic, Baroclinic and Primitive-Equation Models |
| Short-, medium-, long-range prediction | Short-, Medium- and Long-Range Prediction |
| Objective analysis, initialization, data assimilation, satellite data | Objective Analysis, Initialization, Data Assimilation and Satellite Data |
How CSIR NET Asks It
This unit is examined in Part C of the Earth Sciences paper. Per the NTA information bulletin for December 2026:
| Part C (Earth Sciences) | |
|---|---|
| Questions given | 80 |
| Questions to attempt | 25 |
| Marks per correct answer | 4 |
| Marks per wrong answer | −1.32 |
| Maximum | 100 marks |
Expect three kinds of question from this unit:
- Short numericals: the CFL time step, an amplification factor after \( n \) steps, the phase-speed ratio of a short wave, an aliased wavelength, a Rossby wave speed or stationary wavelength, error-doubling arithmetic, a contingency-table score, a Cressman or optimal-interpolation analysis.
- Concept identification: which approximation filters which wave, which variables are prognostic, which initialization sets which quantity to zero, which assimilation method needs an adjoint, which satellite band sounds temperature through cloud.
- Statement sets: “which of the following are correct” about time schemes, model hierarchies, predictability or assimilation. Exact conditions (\( \lvert p\rvert \le 1 \), tendencies versus amplitudes, first versus second kind) decide them.
There is no official breakdown of question counts by unit; nothing on these pages should be read as one.
Quick Reference
| Concept | Formula or fact |
|---|---|
| CFL | \( c_{\max}\Delta t/\Delta x \le 1 \) (1-D) |
| Forward scheme (oscillation) | \( \lvert\lambda\rvert = \sqrt{1 + p^2} \): always unstable |
| Leapfrog | neutral for \( \lvert p\rvert \le 1 \); computational mode; Robert–Asselin filter |
| Centred phase speed | \( c^{*}/c = \sin(k\Delta x)/(k\Delta x) \) |
| Aliasing | \( k^{*} = 2\pi/\Delta x - k \) |
| Rossby–Haurwitz | \( c = U - \beta/(k^2 + l^2) \) |
| PE prognostic | \( u, v, T, q, p_s \) |
| Forecast ranges | short ≤ 3 d; medium 3–10 d; extended 10–30 d; long 30 d–2 yr |
| Error growth | gain \( \tau_d\log_2(\text{reduction factor}) \) |
| Optimal interpolation | weight \( \sigma_b^2/(\sigma_b^2 + \sigma_o^2) \) |
| Cressman | \( (R^2 - r^2)/(R^2 + r^2) \) |
| NNMI | gravity-mode tendencies = 0 |
Related units: Physical Meteorology · Climatology · Numerical Analysis and Inversion · Earth Sciences overview
Found an error? See About for how to report it.
Start Learning
Begin with the first topic: