Instructor: Prof. Ryan Torn (442.4560)

 
Course Objective: This course will describe the components of a basic numerical weather prediction model,
including numerical methods of solving PDEs, how to deal with unresolved features, model initialization and ensembles.

 
Prerequisites: ATM 500, familiarity with partial differential equations and computer programming.

 
Class meets: TR 9:00-10:20 in ETEC 482

 
Syllabus

 
Links

Stochastic Pattern

ECMWF Observation Impact US Model Observation Impact

EnKF Figure 1 EnKF Figure 2

Covariance Comparison

Prior Distribution, Prior+Obs. Distribution, Posterior Distribution, Gaussian Distributions

Observation Network (ECMWF)

Lorenz Model

AI/ML Bias Correction: Neural Network Bias Correction, Results (Zanetta et al. 2023)

Radiation Emulator: Shortwave, Longwave (Lagerquist et al. 2023)

AIFS: 500 hPa anomaly correlation, Tropical Cyclones (Lang et al. 2024)

Data-Driven Models: GraphCast Schematic (Lam et al. 2024), Equator Heating, Rossby Wave, Geo. Adjustment (Hakim and Masanam 2024)

Fine-scale Surface Fluxes

Input Data: Land Use (Fei Chen, NSF-NCAR)

NOAH LSM: Energy, Water, Carbon (He et al. 2023)

Satellite Image

ECMWF RRTMG: Schematic, LW Bands, SW Bands (ECMWF Documentation)

Correlated K

Longwave Absorption Spectra

Plank's Law Radiation

Thompson Scheme: Equations, Aerosols,

Microphysics Schemes: Cloud Drop Sizes (McDonald 1958), Ice Habits, Marshall-Palmer Distribution, Microphysics Flow Chart

Positive Definite Advection (Skamarock and Weisman 2009)

Spherical Harmonics

Effective Resolution: Observed Energy Spectra, Model Energy Spectra, Model Resolution, MPAS/FV3 Resolution

Vertical Coordinate: MPAS Vertical

FV3 Model: FV3 Cell Structure (FV3 Description)

MPAS Model: MPAS Grid, MPAS Cell Structure (MPAS Website)

Leapfrog Time, Centered Space: 20 dx wave, 2 dx wave

Forward Time, Backward Space: 20 dx wave, 2 dx wave

Barotropic Vorticity Model: Example 1, Example 2

Three Stage Methods: Leapfrog (unstable), 2nd Order Adams-Bashforth

Two Stage Methods: 2nd Order Runge-Kutta, Matsuno

Single Stage Time Stepping: Forward Differencing (half time step), Backward Differencing, Trapezoidal

Wave Demonstration: Basic solution, Half dt, dx, Triple dt

Horizontal Derivatives: Forward Differencing, Centered Differencing (double resolution), 4th Order Differencing

Introduction

NWP History Slides

Anomaly Correlation History (ECMWF)

Global Ensemble

Observation Network (ECMWF)

ECMWF Charts

GraphCast Schematic (Lam et al. 2024)

Model Process/Parameterization (COMET Program)

Global Satellite Image

MPAS 3 km Movie (NCAR-NSF)

Earth Visualization

GEOS 3 km Movie (NASA Goddard)