WEBVTT

00:00:00.000 --> 00:00:08.000
Ask a question. Build the experiment.
A suite of simulations, from molecules to planetary dynamics.

00:00:08.000 --> 00:00:14.000
The hard part is everything in between.
Methods. Models. Compute. Diagnostics. Interpretation.

00:00:14.000 --> 00:00:21.000
One more molecule. A different pathway.
Compare the mechanism before committing to the next experiment.

00:00:21.000 --> 00:00:27.000
A surface. Two explanations.
Separate the chemistry of a coating from the geometry of a powder bed.

00:00:27.000 --> 00:00:34.000
Scattered evidence. A testable recipe.
Bring published measurements and a process model into one investigation.

00:00:34.000 --> 00:00:41.000
Then test the moving pieces.
Explore how vibration changes contacts between model powder clumps.

00:00:41.000 --> 00:00:50.000
Search the space. Find the tradeoff.
Screen millions of model candidates, then inspect the feasible designs.

00:00:50.000 --> 00:00:59.000
Change the slope. Watch the response.
Run four terrain scenarios and compare the simulated fire behavior.

00:00:59.000 --> 00:01:07.000
One solver. Very different physics.
Fluid instability. A heated stellar-layer toy. The first instant of an impact.

00:01:07.000 --> 00:01:15.000
A ripple grows into model weather.
Run the dynamics, then explain the mechanism behind the motion.

00:01:15.000 --> 00:01:23.000
Follow the dynamics. Test the explanation.
A month of model weather. A separate study of how a storm splits.

00:01:23.000 --> 00:01:31.000
Separate the cause from the response.
Hold the temperature fixed. Then let the model ocean warm.

00:01:31.000 --> 00:01:39.000
A good fit can hide a different world.
Compare plausible subsurface models against the same synthetic observations.

00:01:39.000 --> 00:01:45.000
Know what the data cannot tell you.
A gravity model can recover bulk properties while leaving depth unresolved.

00:01:45.000 --> 00:01:52.000
Keep the evidence. Keep the uncertainty.
Some runs explain a mechanism. Others show where the answer is not ready.

00:01:52.000 --> 00:02:00.000
From a research question to a suite of simulations.
Models. Runs. Comparisons. Evidence for the next decision.