ROOISPOOK: BUILT OUT OF BAD NIGHTS

ROOISPOOK-PROTOKOL: BUILT OUT OF BAD NIGHTS

Wrong roads, dead horizons, ironstone, Gemini arguments, lightning, mosquitoes and the accidental birth of a South African TLE field protocol.


There should probably be a respectable origin story for Rooispook.

Something involving notebooks.

Equations.

A calm discussion about atmospheric electricity.

Maybe somebody staring thoughtfully at a weather map before saying:

There must be a better way.

That is not what happened.

Rooispook-Protokol was not designed.

It survived.

It survived wrong roads, dead horizons, busted tyres, camera-clock problems, compass arguments, failed captures, city lights, overconfident AI answers, and nights where another observer captured the thing I had driven hundreds of kilometres to photograph while my own camera recorded the statistical properties of empty sky.

There was also quite a lot of profanity.

And mosquitoes.

So many fucking mosquitoes.

By the time Rooispook became something I could honestly call a protocol, most of its useful rules had already been paid for in frustration.

Before Rooispook-protokol: Point Camera at Storm and Pray

Before there was a protocol, there was a camera.

And somewhere beyond the horizon, hopefully, a thunderstorm.

The early method was basically:

Find storm. Point camera. Start intervalometer. Hope.

Obviously there was more technique involved than that, but looking back now, far too much depended on instinct.

The storm looked promising.

Lightning looked interesting.

The horizon seemed roughly correct.

So I would let the camera hammer away into the darkness.

Hundreds of frames.

Thousands.

Nothing.

Another exposure.

Nothing.

Another perfectly focused photograph of stars.

After enough nights like that, you become exceptionally skilled at documenting absolutely fuck-all.

At first, a blank night was just failure.

No sprite.

No useful photograph.

Waste of time.

It took a long time before I realized that the failure itself was trying to tell me something.

THE FIVE-HOUR DRIVE THAT TOOK ALL FUCKING DAY

The first time I went to meet Albertus Horn at Kuruman Hills, the trip should have been straightforward.

Welkom.

Bloemfontein.

Kimberley.

Danielskuil.

Kuruman.

About five hours.

A normal road trip.

Then navigation decided otherwise.

Somewhere along the journey I ended up redirected onto back roads on the Vryburg side.

Then more back roads.

Then roads that appeared to regard tar as an optional design philosophy.

The sun was shining directly into my face.

I could barely see jack shit.

At one stage I was creeping along at something around 20 km/h on a road supposedly meant for considerably more.

The navigation system remained wonderfully confident.

Of course it did.

Turn here.

Continue.

Recalculate.

Go there.

Meanwhile I was developing entirely new branches of the English language.

There were sections where I had to reverse the bloody car.

Places where getting back onto the road apparently required temporarily leaving the road.

I had left Welkom sometime around two in the afternoon.

I reached Albertus somewhere around one in the morning.

A five-hour journey had somehow become an expedition.

And yet something strange happened.

While the navigation was destroying my will to live, the atmosphere was putting on a show.

Convective towers.

Storm cells.

Huge systems building out across the landscape.

Lightning firing in sectors I would never normally have seen.

That disastrous trip showed me parts of the weather I would otherwise have driven straight past.

FIELD RULE EARNED THE HARD WAY:
A journey can go completely wrong and still produce useful observations.

KURUMAN EATS TYRES

Then I met the road surface around the hills.

Ironstone.

Sharp slate-like rock.

Enough exposed material to make every tyre on the vehicle reconsider its career.

And naturally:

flat tyre.

First one.

Lesson learned.

The night before, I had driven sections of that terrain with much more confidence than intelligence.

Then daylight came.

I looked properly at what I had been driving across.

And my remaining tyres collectively delivered a clear scientific recommendation:

Slow the fuck down.

Some of those roads are much more comfortable under a proper 4x4.

I was not driving one.

So the method changed.

Creep.

Choose the line.

Watch the rocks.

Keep the car intact.

FIELD RULE EARNED THE HARD WAY:
Confidence does not change the terrain.

That lesson later turned out to apply equally well to atmospheric data.

NO SIGNAL. NO LIGHTS. NO MOZZIES. PARADISE.

I Have arrived -  Kuruman Hills became something completely different.

No useful internet.

Very little signal.

No city electricity lighting the entire horizon.

No urban glow swallowing the sky.

Just darkness.

Proper darkness.

The Milky Way was no longer a faint decorative stripe above a town.

It became architecture.

You could stand outside after midnight and the sky suddenly had depth.

Far away on the horizon there might be one glow.

A town?

A farm?

A neighbour?

Who knows.

And then I discovered something miraculous.

There were no fucking mosquitoes.

None.


I could not believe it.

I had accepted the terrible road.

The isolation.

The tyre.

No signal.

Snakes.

Darkness.

All of it.

But this?

This was heaven.

I could stand outside late at night with a camera pointed at the sky without slapping my neck every thirty seconds.

Puff adder?

Fine.

Complete isolation?

Fine.

No mosquitoes?

Booyah.

Kuruman had passed the observational-site test that mattered most to me personally.

THE FIRST RED GHOST ON A PHONE

The irony is that one of the breakthroughs came after returning home.

Late October.

Samsung S25.

Hyperlapse.

A storm sector out toward the Natal/Bethlehem side.

And then:

There it was.

My first sprite capture.

On a phone.

Not a giant telescope.

Not a purpose-built scientific instrument.

A cellphone.

That changes the question immediately.

It was no longer:

Can I capture one?

Apparently yes.

Now it became:

How do I stop depending almost completely on luck?

That is where obsession starts becoming engineering.

9 DECEMBER: DID WE JUST SHOOT THE SAME FUCKING SPRITE?

December became even stranger.

I was getting more sprite-like events.

Then Albertus asked a question that suddenly changed the scale of the whole problem:

Did you capture the same sprite?

My response was approximately:

I don't know, dude.

So we exchanged pictures through the storm-chasing group.

His image.

Mine.

Compare structure.

Compare timing.

Compare the sky.

And then came one of those moments where the room becomes very small and the idea becomes very big.


Bloody hell.

There it was.

Potentially the same TLE observed from two widely separated locations.

Now we had something completely different.

Two viewpoints.

A long baseline.

Potential parallax.

The possibility of triangulation.

For a while, the analysis got extremely excited.

Perhaps too excited.

The later review was much more disciplined.

The station baseline was real and useful, but the available files and timing did not support claims of completed triangulation, exact sprite altitude, exact parent lightning or an exact geographic position.

That distinction became one of the most important lessons in Rooispook.

FIELD RULE EARNED THE HARD WAY:
Being excited about the data does not create additional data.

THE GOLDFIELDS COMPASS CIVIL WAR

Then there is aiming.

People who do not photograph distant storms imagine this process as:

“Face southwest.”

Done.

No.

Especially not in the Goldfields.

Windy shows one reference.

Google Maps gives another visual interpretation.

Stellarium gives you the astronomical sky.

The phone compass changes.

The analogue compass has its own opinion.

Move the phone.

Different.

Stand near metal.

Different.

Recalibrate.

Different again.

When your storm is hundreds of kilometres away, a small angular error is not small anymore.

A few degrees can mean aiming at an entirely different portion of distant sky.

Albertus knows the stars well.

Open Stellarium.

Use the sky itself.

Elegant.

Meanwhile I am in Welkom conducting peace negotiations between four navigation systems that refuse to acknowledge one another's sovereignty.

This problem is still part of the journey.

Plate solving should eventually make the camera frame itself an astronomical reference.

Getting that workflow running properly is another bloody project.

It will get there when it gets there.

ROUND TWO: ALBERTUS 1, GÉNE 0

Then I went back to Kuruman.

Second major attempt.

More experience.

Better understanding.

More preparation.

Surely this time...

Albertus captured sprites.

I captured:

fuck all.

Again.

That one nearly finished me.

The first failure can motivate you.

The second failure, after you believe you have already learned the lesson, feels personal.

I was close to:

Fuck this. I'm done. Take the camera. Take the card. Take the whole bloody thing.

You can only chase a millisecond phenomenon for so long before part of your brain starts negotiating surrender.

Albertus gets one.

You don't.

Again.

Wonderful for him.

Psychological warfare for you.


Then the same question crawls back into your head:

What is different?

Position?

Elevation?

Optics?

Framing?

Timing?

Storm sector?

What did his setup know that mine did not?

Every time that question returned, Rooispook grew another bone.

THE GATE INCIDENT

Albertus is also fucking crazy.

I mean that affectionately.

There was one trip where the lightning around us was not decorative.

It was close enough that I had no interest whatsoever in getting out of the car.

I considered this a very sound scientific decision.

Albertus apparently disagreed.

There was this enormous metal gate.

The kind of farm gate that looks capable of regulating giraffe migration.

Lightning around the area.

Ironstone terrain.

Metal fencing.

Everything about the scene suggesting:

Maybe don't become electrically interesting right now.

Albertus gets out.

Goes to the gate.

Flash.

Then:

“WAAAH!”

And suddenly this man is doing a very rapid dance back toward the car.

Me:

I told you, dude. You're fucking crazy.

Then, because apparently one close-lightning lesson is insufficient, the gate eventually still needs closing.

Out he goes again.

Lightning again.

At that stage meteorology had ended.

There were simply two men making questionable decisions around conductive agricultural infrastructure.

For the avoidance of doubt: do not copy this.

Close lightning and large metal structures are not a field technique.

FIELD RULE EARNED THE HARD WAY:
No photograph is worth becoming part of the electrical circuit.

GEMINI, WHERE THE FUCK IS THE STORM?

Gemini deserves a chapter because it genuinely helped start this entire mess.


It was useful.

Sometimes extremely useful.

Storm geometry.

Sprite physics.

Distance.

Lightning concepts.

All of that became easier to explore conversationally.

But Gemini also had a remarkable ability to take an assumption, put it into a beautifully technical paragraph, and return it sounding like reality had already signed the paperwork.

Then I would walk outside.

Look where it said.

Nothing.

Sometimes not even a useful storm.

This created some memorable conversations.

Me:

Where do I aim?

AI:

Over there, essentially.

Me:

Looks over there.

Stars.

Haze.

Streetlight.

No storm.

Me:

WHERE IS THE FUCKING STORM???


That may genuinely be part of the origin of Rooispook.

Not because AI was always correct.

Because it was useful enough to continue using, but wrong often enough that I became increasingly obsessed with independent verification.

Eventually I stopped wanting an oracle.

I wanted instrumentation.

Give me the calculation.

Give me the assumptions.

Give me the likely sector.

Tell me what would invalidate it.

Then allow the human standing outside to make the final call.

FIELD RULE EARNED THE HARD WAY:
Never trust the narrative more than the sky.

WHEN FAILURE FINALLY BECAME DATA

This was the biggest transformation.

Initially:

No sprite = failed night.

Eventually:

No sprite = why?

Wrong azimuth?

Log it.

Bad horizon?

Log it.

Cloud interference?

Log it.

Storm died before the observation window?

Log it.

Timing uncertain?

Log it.

Camera problem?

Log it.

Everything apparently perfect and still no TLE?

Definitely log it.

A blank frame stopped meaning nothing.

It became a calibration point.

Once that happened, Rooispook-protokol stopped being simply a sequence of wins and losses.

It started becoming a dataset.

THE TWO NIGHTS AFTER SUCCESS

(Thank you Albertus Horn)

One of the later field tests makes this particularly clear.

A targeting instruction went out before an observation.

A sprite was then reported within the suggested time window and target sector.

Success.

Excellent.

Then came two nights of zero captures.

Those nights matter too.

If I preserve only the successful prediction and conveniently forget the misses, I have built marketing.

Not methodology.

The misses stay.

They belong in the record.

THE DAY ROOISPOOK STOPPED BEING A HUNT AND BECAME A METHOD

Nobody woke up one morning and announced:

“Today I will invent Rooispook.”

It accumulated.

Wrong directions produced azimuth discipline.

Bad camera clocks produced timing discipline.

Failed captures produced negative-observation logging.

Double-station work produced synchronization requirements.

Missing originals produced file-preservation rules.

Overconfident analysis produced evidence labels.

Streetlights produced shielding.

Kuruman produced geometry.

And frustration produced skepticism.

Eventually the entire mess could be reduced to three lines:

Calculate first.
Observe second.
Verify afterward.

That sounds obvious now.

It was not obvious while standing underneath a streetlight screaming at a phone.

THANKS, GEMINI GEMS. ALSO, FOR THE NEW DISCOVERED PROFANITY!

This is probably the fairest conclusion.

Gemini helped.

A lot.

Without those conversations, Rooispook probably would not have developed in exactly the same way.

Ideas emerged.

Questions became more technical.

Metadata started mattering more.

Geometry started mattering more.

And I gained something available at extremely antisocial hours that I could argue with.

It also occasionally produced a spectacularly confident explanation that did not survive contact with the real atmosphere.

That turned out to be useful too.

Every time an answer failed against the actual sky, I trusted eloquence slightly less and evidence slightly more.

That is not a bad education.

WOULD I DO IT AGAIN?

Absolutely.

Give me Kuruman again.

Give me the terrible road.

Give me the ironstone.

Give me the flat tyre.

Give me the Milky Way.

Give me the ridiculous darkness.

Give me Albertus doing something questionable near a gate.

Give me the camera.

Give me a distant MCS quietly doing something enormous beyond the horizon.

You may keep the mosquitoes.

And somewhere 200 or 300 kilometres away, let the atmosphere do something violent enough to make the mesosphere flash red for a few milliseconds.

I'll go.

Because underneath the software, maps, calculations, photography and swearing, this is still the reason I love it.

You are outside.

The atmosphere is doing something enormous.

You do not completely understand it yet.

And you get to try.

Again.

And again.

And again.

Until the sky finally gives something back.


Calculate first. Observe second. Verify afterward.

And when all three go sideways:

Swear, log the bastard properly, and figure out why.

— Géne C. Morrison
StormJagterVTX / Rooispook Protocol

Comments

Popular Posts