Part 008 · Chapter 4: The First Questions | 1 of 2

CHAPTER 4:
THE FIRST QUESTIONS

The event reached Pretoria as a line of bad data.

Thuli Mokoena was comparing timing correc­tions from two ground stations when a pale vertical mark appeared across the lower third of her display. It was thin enough to miss at normal scale. The automated review had already placed it in the inter­fer­ence queue with a confidence score high enough that no human being was required to look at it.

She looked anyway.

The mark began at 14:10 UTC and lasted a little more than half a second. It appeared across several frequen­cies monitored by the Harte­beesthoek station, strongest in bands used for deep-space support and weakly present in channels where there should have been almost nothing.

Thuli enlarged the raw trace.

The event rose in four uneven steps. Each step occupied a different part of the spectrum. The final drop was abrupt.

She checked the clas­si­fi­ca­tion.

PROBABLE LOCAL TRANSMISSION.

The system had chosen the most common answer. A vehicle near the station. A main­te­nance radio. A satellite crossing the receiver sidelobes. Some piece of human equipment leaking energy into a place where it had no permission to be.

Most anomalies were human.

Thuli opened the station activity log.

No main­te­nance trans­mis­sion. No aircraft inside the protected zone. No scheduled cali­bra­tion burst. The nearest commercial satellite had been below the usable horizon. A catalogued object from a failed lunar launch had crossed the eastern sky, but eleven minutes later and nowhere near the antenna pointing.

She ran the trace against the terres­trial-inter­fer­ence library.

Nothing matched.

The office around her remained quiet. SANSA’s Pretoria coor­di­na­tion room held fewer people than usual because the main afternoon support period had ended. Two analysts worked at the far console bank. A wall display showed solar conditions, orbital traffic and the status of partner stations across southern Africa. The Sun was marked green. Geomag­netic conditions were green. Lightning activity over the interior was low.

Thuli opened the space-weather feeds from Hermanus.

A magnetic-field monitor had recorded a distur­bance at the same second.

Not the same waveform. It could not have recorded the same waveform. The instrument measured a different physical quantity. But the onset aligned within the timing uncer­tainty.

She sat forward.

Harte­beesthoek and Hermanus did not share a receiver chain. They did not share local power. Their clocks were disci­plined through different systems and verified inde­pen­dently. A common software error could still contam­i­nate the central archive, but the raw station buffers would show whether the event existed before collection.

Thuli requested both.

The first buffer arrived imme­di­ately. The second stalled behind an automated storage check.

She called Hermanus.

A night-shift technician answered after three tones.

“I need the unpro­cessed magnetic data from fourteen ten,” Thuli said.

“Today?”

“Yes.”

“What interval?”

“Fourteen oh-nine fifty-eight to fourteen ten oh-three. Full time resolution.”

The technician was silent for a moment. “That is not in the normal archive product.”

“I know.”

“Space weather request?”

“Signal integrity.”

“That is not an answer.”

“It is the answer I have.”

He sighed softly. “Give me five minutes.”

Thuli ended the call and returned to the radio trace.

She removed station correc­tions, then compared the event across individual receiver paths. Two channels showed the first rise. A third joined seven millisec­onds later. A fourth seemed to lead the others, which should not have been possible given the hardware path.

She checked the clock labels.

One channel had been routed through an older digitizer whose internal time stamp was known to drift. That explained part of it.

Not all.

“Are you still on the cali­bra­tion report?” her section chief asked.

He had approached without her hearing him. He carried a tablet under one arm and an empty cup in the other.

“No.”

“That sounded promising.”

She moved the trace to the shared side display. “Harte­beesthoek recorded this at fourteen ten. The classifier called it local inter­fer­ence.”

He examined the waveform.

“And?”

“No local source. No scheduled trans­mis­sion. Satellite geometry does not fit.”

“Receiver fault?”

“Present on separate chains.”

“Clock event?”

“I’m checking.”

He pointed at the four rises. “It looks switched.”

“That is what bothers me.”

“Human systems switch.”

“Yes.”

“Natural systems do not care about our expec­ta­tion of smooth curves.”

“I know.”

He set the cup beside her keyboard. “Solar?”

“Quiet.”

“Lightning?”

“No match.”

“Aircraft?”

“None in the zone.”

“Then leave it unresolved.”

“The magnetic station saw an onset at the same time.”

That held him for a second.

“Same shape?”

“No. Same boundary.”

“Central archive contam­i­na­tion?”

“The Hermanus raw buffer will tell us.”

Her section chief watched the mark on the screen. “How strong?”

“Barely above threshold.”

“Could it affect a spacecraft?”

She turned toward him. “I did not say that.”

“You were thinking it.”

“I was thinking about why it appeared in different instru­ments.”

He glanced at the wall display. Perseus was still shown in green among the deep-space mission tracks. Its public position marker advanced from prediction, but nothing on the SANSA display indicated a problem.

“What was Perseus doing at fourteen ten?” he asked.

“Mid-course correction.”

“You checked.”

“Yes.”

“Did we support the maneuver?”

“Not directly. Harte­beesthoek was monitoring network spill and timing references.”

“Then the alignment may mean nothing.”

“It may.”

The Hermanus buffer arrived.

Thuli opened the file without applying central timing correc­tions. The distur­bance was present in the station’s local record. A small deflection in two axes, then a return. It was not large enough to trigger a space-weather notice.

She compared the local clock against the daily reference test.

The clock had remained stable.

Her section chief leaned closer.

“That still does not connect it to the radio event.”

“No.”

“Or Perseus.”

“No.”

“Good.”

He said the word as if she had passed a test.

Thuli reduced the trace to a statis­tical comparison. The chance alignment was not remarkable when measured against every small distur­bance the network recorded in a day. That was the problem with weak events. Enough instru­ments produced enough noise that any chosen second could be made to look important after the fact.

She did not want to make it important.

She wanted to know why it had edges.

A secure noti­fi­ca­tion appeared at the top of the wall display.

PRESERVE RAW NETWORK DATA: PERSEUS COMMUNICATIONS ANOMALY.

The room changed.

One of the analysts at the far bank stood. The other stopped speaking in the middle of a sentence.

Thuli opened the notice.

It contained no descrip­tion beyond the preser­va­tion order and a request for station logs beginning at 14:09 UTC. The distri­bu­tion list included NASA, ESA, Lunar Gate operations, several national agencies and the UN-linked planetary coor­di­na­tion channel.

The requested interval covered her event.

Her section chief read over her shoulder.

“Freeze the files,” he said.

“I already have them.”

“Freeze everything around them. Receiver config­u­ra­tion, clocks, envi­ron­mental logs, satellite catalogue, main­te­nance channels.”

Thuli began issuing archive locks.

“Do we send the anomaly note?” she asked.

“We send the requested raw data.”

“With the classifier label attached?”

He looked at her.

“The classifier says probable local trans­mis­sion.”

“And you disagree?”

“I do not know enough to agree.”

“Then change it to unresolved.”

“That will not tell them why.”

“They have their own analysts.”

“That is not a reason to remove ours.”

“It is a reason not to attach a theory to a weak trace during a spacecraft emergency.”

“I am not proposing a theory.”

“You are proposing relevance.”

“The time makes it relevant.”

“The time makes it dangerous.”

He spoke quietly. Nobody else in the room could hear them.

Thuli looked again at the preser­va­tion notice. It did not say Perseus was lost. It did not say the crew were in danger. It said commu­ni­ca­tions anomaly, the same language used for missed handovers and antenna faults.

The requested data window began one minute before the trace.

Somebody in mission control considered the timing important enough to ask the world for raw records.

“I want a technical note attached,” she said.

“Internal.”

“Why?”

“Because if it is nothing, we do not need a South African agency appearing to claim an unex­plained pulse when eight people are missing.”

The number reached her before the rest of the sentence.

“Missing?”

He did not answer imme­di­ately.

The secure notice on his tablet carried a higher access level than the one on the wall. Thuli saw the clas­si­fi­ca­tion banner before he lowered it.

“No channels have responded,” he said. “That is all we know.”

“How long?”

“Long enough to preserve the data.”

Thuli looked at the blank green status beside Perseus.

The public display had not changed.

“What do they have from other stations?” she asked.

“I don’t know.”

“Have they reported anything like this?”

“I don’t know.”

“Then how do we know ours is weak?”

“Because it is weak.”

“That is not the same question.”

Her section chief picked up his empty cup.

“Write the note. Keep it factual. No causal language, no spec­u­la­tion, no words like artificial, attack or message. I will review it before it leaves SANSA.”

“Where does it leave SANSA?”

“Through the national channel.”

“Only?”

“For now.”

He walked back toward his office.

Thuli watched him go, then opened a new report.

She entered the time, duration, instru­ments and confidence limits. She described the radio waveform as structured only in the narrow sense that its rise was not random inside the observed interval. She wrote that the magnetic distur­bance was temporally aligned but not demon­strably related. She listed every ordinary expla­na­tion she had tested and every one she had not yet excluded.

At the end, she wrote:

NO CLAIM OF TRANSMISSION OR ARTIFICIAL ORIGIN IS SUPPORTED.

She read the sentence twice.

Then she added:

THE EVENT SHOULD NOT BE DISCARDED AS LOCAL INTERFERENCE WITHOUT INDEPENDENT REVIEW.

At 18:37 UTC, the public mission page changed.

MISSION PERSEUS EXPE­RI­ENCED AN INTER­RUP­TION IN SCHEDULED COMMU­NI­CA­TIONS FOLLOWING A SUCCESSFUL ROUTINE COURSE CORRECTION. GROUND TEAMS ARE WORKING THROUGH ESTAB­LISHED PROCEDURES TO REESTAB­LISH CONTACT.

The analysts gathered around the wall display.

Nobody said what the statement left out.

Thuli checked the public tracker. Perseus remained on its predicted line. Under commu­ni­ca­tions, UPDATE PENDING had replaced nominal.

Her technical note waited in her section chief’s review queue.

She opened the distri­bu­tion options.

The national channel was selected. The UN-linked scientific liaison was not. Neither was the inter-obser­va­tory veri­fi­ca­tion route shared with South African radio astronomy partners.

Thuli moved the cursor over the second option.

Her section chief had told her where the report should go. He had not told her she could not ask another scientist to verify the instrument behavior. Inde­pen­dent validation was part of the anomaly protocol. It was intended for cali­bra­tion disputes, not missing spacecraft, but the procedure did not distin­guish between comfort­able and uncom­fort­able data.

She selected the radio astronomy veri­fi­ca­tion route and addressed the blind packet to Anika Mwale.

The packet contained no Perseus reference. Only raw timing, station geometry, the waveform, and a request:

CHECK FOR CORRELATED TRANSIENT BETWEEN 14:09:59 AND 14:10:02 UTC. DO NOT USE CENTRAL ARCHIVE PRODUCT.

Thuli authen­ti­cated the request.

Before sending, she added the file hashes for the original buffers.

Then she released it.