NOADS in 10 minutes

NOADS in 10 minutes#

This guide walks you through a complete optimization — from setting up an aviation decarbonization scenario to interpreting the results.

Prerequisites

Make sure you have installed NOADS and are familiar with the core concepts.

Step 1: Run a single-policy optimization#

The fastest way to get results is through the high-level API. This sets up a complete scenario (fleet, energy system, traffic demand, constraints) and runs the optimizer:

from noads.application.examples import single_policy_scenario_optimization

output = single_policy_scenario_optimization(
    global_scenario_name="SSP2-26",   # (1)!
    technology_index=1,                # (2)!
    carbon_budget_percent=3.0,         # (3)!
    drop_in_only=False,                # (4)!
    fossil_kerosene_only=False,        # (5)!
    plot_optimum=True,                 # (6)!
    save_optimum=True,                 # (7)!
)

  1. Climate scenario: SSP2 pathway with 2.6 W/m² forcing (≈2°C target).

  2. Technology level: 0 = pessimistic, 1 = mid, 2 = optimistic assumptions for battery density, motor power, structural efficiency, etc.

  3. Carbon budget share: aviation gets 3% of the global remaining CO₂ budget.

  4. Drop-in only: if True, restricts to kerosene-compatible fuels (no hydrogen, no batteries).

  5. Fossil kerosene only: if True, disables all alternative fuels (baseline scenario).

  6. Plot: automatically generates result plots.

  7. Save: writes the optimal solution to disk for later reuse.

The optimizer (NLOPT SLSQP with JAX-computed gradients) typically converges in a few minutes. You will see an optimization history and result plots showing emissions, fleet composition, and energy mix trajectories from 2025 to 2075.

Step 2: Understand the scenario setup#

Behind the scenes, single_policy_scenario_optimization calls single_scenario_setup, which you can also use directly for more control:

from noads.application.scenario_setup import single_scenario_setup

scenario, design_space, constraints, energy_mix, fleet = single_scenario_setup(
    name="my_scenario",
    background_scenario_name="SSP2-26",
    start_year=2025,
    end_year=2075,
    technology_index=1,
)

This returns five objects:

Object

Type

What it contains

scenario

TemporalScenario

The assembled time-dependent model (fleet + energy + traffic)

design_space

GEMSEO DesignSpace

Decision variables: fuel blend shares, aircraft entry dates, market shares

constraints

dict

Inequality constraints: carbon budget, resource limits, feasibility bounds

energy_mix

EnergyMix

The energy production system (pathways, carriers, impacts)

fleet

Fleet

The aircraft fleet (current + new aircraft, market segments)

Step 3: Run the optimizer with GEMSEO#

With these objects, you have full control over the optimization:

from gemseo import create_scenario

gemseo_scenario = create_scenario(
    disciplines=[scenario.discipline],
    formulation_name="DisciplinaryOpt",
    objective_name="cumulative.CO2",
    design_space=design_space,
)

# Add constraints
for name, (value, positive) in constraints.items():
    gemseo_scenario.add_constraint(name, "ineq", value=value, positive=positive)

# Run
gemseo_scenario.execute(
    algo_name="NLOPT_SLSQP",
    max_iter=2000,
    ftol_rel=1e-15,
    ineq_tolerance=1e-4,
)

# Retrieve optimal outputs
x_opt = gemseo_scenario.optimization_result.x_opt_as_dict
output = scenario.discipline.execute(x_opt)

Step 4: Visualize results#

from noads.application.visualization import plot_single_scenario_result

plot_single_scenario_result(
    scenario_name="SSP2-26",
    output_optimal={**x_opt, **output},
    energy_mix=energy_mix,
    fleet=fleet,
    low_demand=False,  # True only for the low-demand (supply-cap) formulation
    save_figs=True,
    directory_path="results/my_scenario",  # created automatically if missing
)

This generates plots for:

  • Emissions: CO₂ trajectory vs. the allocated carbon budget

  • Fleet composition: market shares of conventional and new aircraft over time

  • Energy mix: evolution of fuel types (fossil kerosene, biofuel, e-fuel, hydrogen)

  • Traffic: RPK demand trajectory with demand avoidance effects

Step 5: Compare scenarios#

To explore how results change across climate futures or technology assumptions, loop over configurations:

import itertools

scenarios = ["SSP1-19", "SSP2-26", "SSP5-45"]
tech_levels = [0, 1, 2]  # lower, mid, upper

for scenario_name, tech in itertools.product(scenarios, tech_levels):
    output = single_policy_scenario_optimization(
        global_scenario_name=scenario_name,
        technology_index=tech,
        save_optimum=True,
        plot_optimum=False,  # plot later in batch
    )

For robust optimization across multiple SSP pathways simultaneously, use the multi-scenario API:

from noads.application.scenario_setup import multi_scenario_setup

multi_scenario, design_space, constraints, energy_mix, fleet = multi_scenario_setup(
    name="robust",
    background_scenario_names=["SSP1-19", "SSP2-26", "SSP5-45"],
    start_year=2025,
    end_year=2075,
    technology_index=1,
)

This finds a single policy that performs well across all three climate futures.

Step 6: Load and replot saved results#

Saved results can be reloaded without re-running the optimization:

output = single_policy_scenario_optimization(
    global_scenario_name="SSP2-26",
    technology_index=1,
    load_optimum=True,   # load from disk
    plot_optimum=True,   # regenerate plots
)

What’s next?#