Solanar Graduate Program✦Track 02 · Numerical Weather Prediction✦Data Assimilation · Ensembles · Hybrid AI/ML✦A++ Graduate Curriculum✦80% Hard Mastery✦Higher Learning Mathematics Integrated✦NAIB Research Mentor✦
NWP/DA Graduate EvidenceComputational experiments · diagnostics · research
Track HomeNumerical LabNAIB Mentor
Transparent computational science

Evidence Forms

Every substantial experiment records configuration, assumptions, uncertainty, provenance, results and reproducibility so model output never becomes an unexplained black box.

Numerical Experiment Design Record

Question, hypothesis, baseline, perturbation, controlled variables, metrics and reproducibility plan.

Open Form

Model Configuration & Provenance Record

Grid/domain, coordinates, time step, dynamical core, physics, initial/boundary conditions, code/data versions.

Open Form

Numerical Stability & Error Analysis

CFL, truncation error, dispersion, diffusion, conservation and sensitivity.

Open Form

Grid & Coordinate Analysis

Horizontal/vertical grid, metric/coordinate implications, resolution and numerical tradeoffs.

Open Form

Physical Parameterization Audit

Radiation, microphysics, PBL, surface, convection and sub-grid assumptions.

Open Form

Deterministic Model Verification

Bias, MAE/RMSE, categorical scores, spatial/timing error and failure diagnosis.

Open Form

DA Observation Inventory & QC

Observation types, representativeness, metadata, QC, thinning/superobbing concepts and exclusions.

Open Form

Innovation / Residual Diagnostics

Background-minus-observation, analysis-minus-observation, bias, variance and suspect-data patterns.

Open Form

Variational DA Analysis

Cost function, control vector, background/observation terms, minimization logic and assumptions.

Open Form

Tangent-Linear / Adjoint Reasoning Log

Perturbation propagation, sensitivities, linearization assumptions and limitations.

Open Form

Kalman / Ensemble DA Experiment

Prior, observation, uncertainty, gain/update, ensemble spread and posterior diagnosis.

Open Form

Error Covariance Audit

Background/observation covariance structure, localization/inflation concepts and sensitivity.

Open Form

Ensemble Prediction & Probability Brief

Perturbations, spread, clusters, threshold probabilities, reliability and scenarios.

Open Form

ML / Hybrid Prediction Evaluation

Training/evaluation separation, baselines, generalization, physical consistency, uncertainty and failure modes.

Open Form

Computational Reproducibility Record

Dataset provenance, configuration, code logic, seeds, versions, outputs and replication check.

Open Form

Graduate Forecast-System Capstone Record

Architecture, equations, numerics, DA, ensemble design, verification, limitations and research contribution.

Open Form