The military-industrial scientific research system of the academic master.

You say that Danzhou is the capital, and it is indeed the capital.

But in fact, it is nearly 100 kilometers away from Beijing University of Aeronautics and Astronautics.

Even after more than 20 years, this place is still considered a place that has not been fully developed.

With this distance, Chang Haonan and Meng Zhenyuan definitely couldn't go back at night and come back the next morning.

So the two of them spent the night in the measurement and control center.

Comrades in the aerospace system are busy without distinguishing between day and night, so there is nothing else in this place, but there is no shortage of folding beds and other things in the empty rooms.

I had originally planned to finish the work the next day and then go back to wait for the results of the on-site survey.

But the time taken to extract data this time far exceeded Chang Haonan's expectations.

It was not until the early morning of the third day that the concept decomposition was almost completed.

This shows that the picture contains far more "useful" information than he had expected.

Fortunately, this algorithm is still in its early stages, so we are done here.

The subsequent feature recognition and selection cannot be done by computer because the data has not been verified yet.

Similar to previous applications in the radar field, the algorithm can only extract signal features with specific differences.

As for what the specific results are for each feature.

At least at the beginning, it requires human calibration.

Only after a sufficient number of calibrations can judgments be made based solely on the characteristics of the identified signal.

It is similar to how sonar determines the model of the target submarine through soundprint information.

Chang Haonan and his colleagues returned to the hall of the measurement and control center, stood in front of the screen, and looked up at the remote sensing data spectrum after dimensionality reduction processing using the manifold learning algorithm.

There are more than one abnormal situation that Li Nianheng discovered before in this picture.

So it shouldn't be an occasional bug.

But it’s not available everywhere.

So it doesn't seem like simple noise interference.

"I think we should first figure out which sensor captured these strange data."

Chang Haonan was the first to break the silence:

"It seems that in the past, this data should have been directly blocked as noise, but after using the new feature extraction method, it was exposed."

Li Nianheng nodded, obviously agreeing with this opinion.

So, he found Ye Jianpei, the deputy chief engineer in charge of the design of the satellite detection system.

Although the probe itself is not entirely domestically produced, the overall design of the entire satellite was completed entirely by China.

So everything is still under control.

“Raw data alone may not be enough.”

After listening to Li Nianheng's description of the situation, Ye Jianpei thought for a moment, then expressed his position cautiously:

“It would be better to have a log of the entire remote sensing data extraction process, so that the reverse deduction would be more efficient.”

"We previously performed the extraction operation on server 02, and the operation log should still be saved on it."

Chang Haonan replied:

“Part of the dimensionality reduction process is a black box even for me, but if it’s just tracing the source, it shouldn’t have much of an impact.”

Ye Jianpei did not answer, but just nodded, then walked quickly towards the server room with two of his technicians.

"It seems... you will have to stay here for another night."

Li Nianheng looked at Ye Jianpei's hurriedly leaving back and said to Chang Haonan and Meng Zhenyuan beside him.

"It doesn't matter."

Chang Haonan waved his hand:

"I also want to see where those abnormal signals come from."

This was a question that Chang Haonan had never encountered before, including in his previous life, and it did make him somewhat curious.

……

"Boss Chang, Boss Chang?"

After an unknown amount of time, Chang Haonan, who was leaning back in a chair and half-closed his eyes to rest, suddenly heard someone calling his name.

He quickly opened his eyes and stood up, draping the blanket that someone had put on him over the backrest.

Li Nianheng and Ye Jianpei were standing not far behind the chair.

"The results are out?"

"Correct."

Ye Jianpei handed a document to Chang Haonan in a concise manner:

"An L-band polarimetric interferometric synthetic aperture radar, which is part of the hyperspectral remote sensing equipment set we imported from France."

"This device was designed to measure target elevation data and obtain detailed parameters of surface cover, such as forests, ice and snow, and houses. But now it is clear that it has also captured some information that we had not expected before."

Chang Haonan took the document and looked at it, and found that it contained detailed data of this L-band radar.

Even in the manual provided by the French, there is no mention of any other functions besides the two above.

Apparently, they don't know either.

"Is it possible... that it's some kind of resource buried underground?"

At this time, Meng Zhenyuan's thoughts have diverged:

"Like oil fields, coal mines, or..."

When he said this, his eyes lit up.

Obviously, if we can really find so many mines in the country, it will definitely alleviate the urgent problem of energy shortage in the past two years.

Even if he only played a supporting role, he could probably get some special allowance.

However, Meng Zhenyuan's fantasy was soon shattered.

Chang Haonan held his forehead with one hand and replied helplessly:

"How is it possible... It's almost the 21st century. The mines that have not been discovered by humans yet are conservatively estimated to be more than a thousand meters deep. We are using electromagnetic waves, not seismic waves. Even the L-band can only penetrate 20 to 50 at most..."

Having said this, Chang Haonan's voice suddenly stopped.

As a decimeter wave radar with a relatively long wavelength, the L band may theoretically have a certain ability to penetrate the ground.

However, it was generally believed in the past that the penetration of electromagnetic waves depended on the composition of the surface covering layer, so it was not very stable and the resolution was also relatively general, making it generally difficult to achieve the effects people wanted.

The very few ground-penetrating radars that have been put into use are also based on the longer P band, taking advantage of the greater penetrating power of long waves to conduct some rough searches.

For example, looking for shallow underground caves.

But judging from the picture they just obtained.

It is true that the penetrating power of electromagnetic waves is not very stable.

But the effect may not be bad.

It is more likely that there were not enough technical means in the past, so a lot of useful feature data was directly blocked as noise.

Thinking of this, Chang Haonan turned around and looked at the other three people:

“Is there such a possibility?”

He pointed to the big screen and said:

"Even if they are the same topographic features, there are still differences inside them that we don't know about, or that could not be observed by remote sensing in the past?"

"what do you mean……"

This sentence was a bit confusing, but Li Nianheng, an expert in remote sensing technology, quickly understood what Chang Haonan meant:

"For example, deserts can be divided into rock deserts, gravel deserts, sand deserts and mud deserts. These four geomorphic features are also different in our past remote sensing data. Do you think these four geomorphic features can be further subdivided?"

"Yes."

Chang Haonan affirmed.

At this time, Meng Zhenyuan next to him also intervened:

"Then we can only ask the comrades from the Exploration Institute to go to the site to collect samples and then do a detailed analysis."

There was obvious excitement in his tone.

Although he didn't find the mine as he thought, if Chang Haonan's guess is true, then...

Well, I can’t think of any practical significance for the moment.

But at least in the field of geology, the names of the four people present today will definitely be recorded.

Moreover, practical significance is often discovered in the process of seemingly useless research.

What if.

But Chang Haonan shook his head:

"The area with abnormal characteristic signals on this satellite image is very wide, not just in Longyuan. The exploration power of the oil system alone may not be enough."

After saying that, he looked at Li Nianheng.

As the person in charge of the remote sensing mission, he will inevitably have to deal with relevant geological units.

In terms of professional research, they are obviously more professional.

He didn't even need to say anything, Li Nianheng already understood what he meant.

The latter immediately called a staff member from behind:

"Xiao Song, contact the land and resources department immediately and ask them to coordinate with the comrades from the Huaxia Geological Survey Bureau to cooperate with us and conduct on-site exploration of these points."

"All...all?"

There were dozens of marked points on the map. The person called over was a young man who looked like he had just started working and had obviously never seen such a situation.

Li Nianheng pondered for a moment, and after calming down, he felt that the noise was a bit too loud, so he chose a compromise:

"If the personnel and funds are really limited, at least we can start with the Loess Plateau in the east of Longyuan and the desert area of ​​Alxa in the west of Inner Mongolia. I think the characteristic signals extracted from these two areas are more complex."

Complexity means a large amount of information.

Since we are going there anyway, we should definitely give priority to places like this.

"In the eastern part of Longyuan, we can ask Huaxia Petroleum's Exploration and Development Research Institute to send personnel to assist them."

Chang Haonan added at the end.

"Yes!"

After answering simply, the staff member turned around and trotted out of the hall.

"Professor Meng."

Chang Haonan let out a long sigh and turned his attention back to the picture displayed on the big screen:

"I'm afraid our project won't be over so soon."