Sevastopol imagery 7 June 2022

Another imagery update of Sevastopol provided by Capella, this time dated 7 June 2022.

Not too many changes but there is one strange occurance.

Overall, most of the Russian navy ships remain the same. On the north side of the bay, a couple of civilian merchant vessels were collecting grain/wheat from the terminal. Project 02690 Floating crane SPK-54150 had been operational on the southern side but was back next to the grain terminal at the time of the collection.

The remaining ships are same as those in the 31 May 2022 update – except one Project 1239 Dergach class had departed on 5 June 2022.

On the south side in Pivdenna Bay, very little change. Project 02690 Floating crane SPK-46150 was present but had been operational – to then depart a few days later on 8 June 2022 (more on this later).

The submarine pen was open and one Kilo class SSK was no longer present. This was to be found in the maintenance bay 2 km northeast of Pivdenna, on the south side of Sevastopol Bay.

Even stranger was that, along with the Capella imagery here, others showed the Kilo balancing on the deck of a small floating crane. @GrangerE04117 on Twitter concluded it was Project 877V Alrosa – which I agree with.

The remaining Kilo in Pivdenna Bay was confirmed later on by @Capt_Navy

Alrosa balancing on the deck of the floating crane in such a way is something I haven’t seen before. There are floating docks available, but these are in use. Moreover, potentially this method is a faster way of carrying out the work they need to do on the Kilo. How they got it up on the deck is another question!

SPK-46150 left at 1205 UTC on 8 June 2022, probably for Snake Island. The Floating crane had two Tor-M on its deck. The last position on S-AIS came in at 1422 UTC, northwest of Sevastopol. It appears to be following the same route SPK-54150 took previously, so at 6 knots would take approximately 22 hours from that position to reach Snake Island. A rough ETA would be 1230 UTC on 9 June 2022 if it isn’t there already.

SPK-46150‘s activities prior to departing Sevastopol

The use of the Floating cranes as a Tor-M delivery method to Snake Island is certainly a strange one. I said on a Twitter thread that it may be a “one ship fits all” reasoning, rather than using small landing craft or other vessels that may then need a crane to lift the SAM systems onto the jetty. I can’t see any other reason why they’d do it. Unless there are issues with using the Serna class ships at the ramp at the harbour?

It’s certainly a big risk. As I said on the thread. It’s just an idea as to why they might be using the floating cranes but “I’m not saying they’re correct in their methods“.

Admiral Kuznetsov dry dock update

**Update to the small update**

**Imagery amendment – The northern floating crane at the dock entrance is actually a fixed one on the wall. – Thanks to Capt(N) for posting an image that shows this**

A few more Capella Space collection passes were tasked after Admiral Kuznetsov was moved to the 35th Shipyard dry dock.

These were dated 26 and 27 May 2022.

They show that work has started again on the dry dock entrance. Here they will likely seal the mouth up with a temporary steel barrier that has been pile driven into the river bed. From that they can then empty the dry dock and construct the full gate system.

Why they didn’t do this at the time of construction is anyone’s guess, but it is likely they wanted Kuznetsov into the dock as soon as possible so that they can continue the work on the ship.

Three floating cranes appear to be back in attendance to help with the work. The image for 27 May looks like a barrier is already in place, but this is the northern crane.

They used this method to construct the dry dock in the first place, but had to destroy it so that Kuznetsov could be floated in.

In theory, they could use the dry dock as soon as it is empty for any work on the hull that would normally be below the waterline, but this could be dangerous. And with the luck Kuznetsov has had recently…. well, anything could happen!

But, the Russian Navy does appear to like risk and I think they’ll put the lower dock to work as soon as they can. Especially if Kuznetsov has been damaged below the waterline in the previous incidents.

Sevastopol Imagery 31 May 2022

An early morning collection by Capella Space of Sevastopol on 31 May 2022 showed that Project 02690 Floating crane SPK-54150 was possibly back at the base. It had recently been spotted at Snake Island in imagery from Maxar and Planet.

It can be confirmed that the crane is certainly not SPK-46150 as this has been operational all day on the south side of Sevastopol bay according to AIS data from FleetMon.

Also present was a single Project 11356M Admiral Grigorovich class FFGH, two Project 1135 Krivak class FFMs and several Project 775 Ropucha class LSTMs.

Two Kilo class SSKs are in the submarine pen, whilst two Project 1239 Dergach class PGGJMs are north side – these are Bora (615) and Samum (616) though identifying which is which is not possible. SPK-46150 was still at its mooring at the time of the pass.

One of the Dergach class was captured on video in the last few days, though again, with no pennant/hull number, it can not be identified.

AIS data from FleetMon shows SPK-46150 has been active on the south side of Sevastopol Bay most of the morning of 31 May 2022

Admiral Kuznetsov on the move – but not far!

With rumours filtering through that Project 1143.5 CVGM Admiral Kuznetsov was due to move sometime between 17 and 19 May 2022 from its “temporary” mooring position in Murmansk to a purpose built dry dock just a little further south, I set up at collection task with Capella Space to catch before and after imagery of the event.

Kuznetsov had ended up at its mooring position after floating dock PD-50, of the 85th shipyard, sank on 30 October 2018 whilst the CVGM was being floated out after a month of works. During the accident, a crane that was part of the dock fell onto the flight deck causing considerable damage.

That wasn’t the end of the woes for the already delayed refit Kuznetsov was undertaking – originally planned to start in 2017 and already a year late. On 12 December 2019 the ship suffered from a major fire that the United Shipbuilding Corporation (USC) estimated would cost 350 million roubles ($4.7 million/£3.7 million at the time) to repair.

An agreement was made with the Russian MoD that two dry docks of the 35th shipyard in Murmansk would be redesigned and knocked through into one large dry dock that could take Kuznetsov and other large Russian navy ships and submarines.

Work commenced on the new dry dock mid to late 2019 and was due to be completed in early 2021 for Kuznetsov to enter and complete the overhaul. Currently, the 35th Shipyard are restricted to works that can take place alongside.

However, the dry dock is still under construction due to several delays in the construction process. This hasn’t deterred the Russian navy from getting Kuznetsov into the dock.

On 20 May 2022, Kuznetsov made the 1.5 km journey with the assistance of tugs rather than under its own power.

Telegram poster, Arctic Observer – Murmansk (Арктический обозреватель – Мурманск) was the first to post imagery of Kuznetsov on the move on 20 May.

They then posted further imagery a little later.

Capella Imagery

The collection request was made to Capella to cover 17 – 20 May. Typically there wasn’t a collection slot available on the 20th, but the request was extended to the next available on the 22nd.

Low resolution EO imagery on Sentinel was only available for 15 May. After this, the region was 100% cloud covered, making further collections of EO imagery impossible. This is where SAR collections from Capella excel – being able to see, no matter the weather.

Sentinel imagery dated 15/5/22 showing Kuznetsov and the “new” dry dock to the south
The dry dock on 17 May 2022. Working is taking place at the entrance. At least two floating cranes are present.
18 May 2022. Work continues on the dry dock.
19 May 2022. Work appears to have been paused at the dry dock and the entrance cleared.
19 May 2022. At Kuznetsov, a possible tug or floating crane is present. No such activity was taking place on the previous days collections.
22 May 2022. Kuznetsov in the dry dock at 35th Shipyard.

FleetMon S-AIS data

The move used at least four tugs according to S-AIS data from FleetMon. These were – Bizon, Grumant, Helius and Kapitan Shebalkin.

FleetMon AIS data shows Tug Bizon alongside Kuznetsov on 20 May 2022.
FleetMon AIS data shows Tug Grumant alongside Kuznetsov on 20 May 2022.
FleetMon track history for Tug Grumant clearly shows it helped with the Kuznetsov move. All the other tugs mentioned also showed similar tracks to this.
FleetMon AIS data shows Tug Helius working at the dry dock on 20 May 2022.
FleetMon AIS data shows Tug Kapitan Shebalkin alongside Kuznetsov and at the dry dock on 20 May 2022.

Since 20 May, further imagery has been published that shows Kuznetsov in the dry dock. bmpd on LiveJournal has some particularly good ones which showed some of the work being carried out.

Imagery posted on bmpd LiveJournal – courtesy of Pavel Lvov / RIA Novosti.

A couple of the images are interesting as they show potential changes to the weapons systems. Below, it can be seen that the RBU-12000 ASW rocket launchers (designed specifically for Kuznetsov) have been retained (central, far left of image) but the AK-630M on the deck balcony below has been removed.

Imagery posted on bmpd LiveJournal – credited to Alexander Loginov, Anna Savicheva, Svyatoslav Ivanov / severpost.ru

The same has taken place on the starboard side of the ship.

Imagery posted on bmpd LiveJournal

A further image on RIA Novosti credited to Pavel Lvov, taken from above also shows the removal of the AK-630Ms along with the eight Kortik/Kashtan CADS-N-1A each fitted with twin AO-18K (6K30GM) 30 mm rotary cannon and eight SA‐N‐11 (9M311) ‘Grison’ missiles.

Imagery from RIA Novosti credited to Pavel Lvov

The Kashtan is likely to be replaced by Pantsir‐M/Pantsir‐SM CIWS hence their removal.

The image above also shows a lot of surface oil. Whether it is from Kuznetsov or the tugs is anyone’s guess – but I have a feeling I know which one it is

Sevastopol imagery – 15 March 2022


  • Two target ships sought after in imagery
  • Both shown to not be in Sevastopol Bay
  • One turned up a day later

I was keen to know the location of two Russian navy ships that were operating in the vicinity of Sevastopol and the Black Sea region.

The first was Project 1164 Slava-class CGHM Moskva. From satellite imagery available on Sentinel, it was known she had arrived on or around 9 March 2022. She was still present in imagery available from 14 March 2022.

Moskva almost always ties up at the same location so is easy to locate when at the base. In Sentinel SAR imagery (and EO for that matter) you can also measure the length to help assist with the ID.

With the events taking place in the Black Sea, I thought 5 days was quite a long time to be at the base, so it was worth seeing if she was still there on the 15th – Sentinel imagery for that day wasn’t available at the time.

My second target was Project 22160 Bykov-class Corvette Vasily Bykov. If you’ve read my previous blogs, I didn’t believe she had been sunk, and even thought she was elsewhere in the the theatre of operations – possibly the Sea of Azov which I had been monitoring since the alleged “sinking”. This operating area is just a guess though. I’m sure we’ll never really find out.

Moreover, there had been rumours that Vasily Bykov was to always work with Moskva so if one was definitely in Sevastopol, based on the “rumours”, they both should be.

I also had a hunch, that if my guess about being elsewhere was correct, then maybe Vasily Bykov could have arrived anyway, regardless of being with Moskva or not. Having been out the same length of time, she must have needed resupplying as much as Moskva did.

I requested an image collection from Capella on the morning of 15 March 2022, and was lucky enough to get a pass that evening at 1826z, about seven hours after the request had gone in.

This revealed that both Moskva had departed, and that Vasily Bykov was not in.

Whilst this might be looked upon as negative, it isn’t. Intel is Intel. It was now known that Moskva was on her way somewhere and had been stocked up – as it turns out in imagery available later in Sentinel, to take part in operations east of Odessa.

It also showed that Vasily Bykov wasn’t operating with Moskva as per the rumours.

And, low and behold on 16 March 2022, Vasily Bykov did turn up at Sevastopol. A miracle one would say, bearing in mind it was supposed to have been sunk a few weeks earlier.

It wasn’t a bad guess she’d turn up – just 24 hours later than my hunch.

The Capella imagery also showed that there wasn’t much else in the bay. The southern area was empty bar one Kilo-class SSK.

The area next to Moskva‘s normal home was also pretty empty. Just one possible Project 1135M Krivak II-class FFM was present. The imagery for this is a little blurred due to the angle of the collection (44 degrees) and the sweep of the SAR itself. This places the ship almost on its side, but the profile does look like a Krivak-II.

If not, it is a Project 11356M Grigorovich-class FFGH – they are the same length, though the profile is slightly different due to the heli-deck.

This doesn’t appear to have the heli-deck and looks to be stepped down to the stern for accommodate the two AK-100 guns.

Regardless, the imagery from Capella was well timed. Whilst the areas out at sea were clear, over Sevastopol itself it was cloudy so EO wasn’t usable – Sentinel didn’t have any EO passes there anyway – and the Sentinel SAR is nowhere near as good as Capella’s.

Unfortunately, I have no collections available to me over Sevastopol today (16 March 2022) so I can’t see Vasily Bykov, and it looks like other ships are also returning – with Project 775 Ropucha-class LST Kondopoga reported to have arrived too.

170+ Military vehicles at Belarus/Ukraine Border 0757z 24/2/22


*Edited Version


  • Capella Space imagery collected at 0757z on 24/2/22 shows 170+ vehicles
  • Located at 51.5136 29.850
  • 5 km NW of Belarus/Ukraine Border
  • 21.5 km NW of Chernobyl

Recent imagery released by Capella Space shows at least 170 military vehicles waiting in multiple columns at an abandoned village in Belarus. It is located 5 km to the NW of the Belarus/Ukraine border, and 21.5 km from Chernobyl.

At least one column is over 700 metres long and is on the main road to the border crossing point.

The image was collected at 0757z on 24 February 2022, hours before combat at the Chernobyl Nuclear Power Station resulted in Russian Forces taking control of the area.

The imagery cannot distinguish vehicle types or to whom they belong though it is presumed Russian military waiting to cross into Ukraine.

For what it’s worth…..


  • Project 677 Lada-class SSK Sankt Peterburg still in Kronstadt
  • Imagery proves early fake news prior to recent events around Ukraine borders

For what it’s worth….. indeed.

This imagery provided by Capella Space was supposed to have been in a Janes article last week, but the events surrounding Ukraine snowballed so quickly, it almost became old news before it had even really been put out there.

Anyway, rather than letting the imagery collection go to waste, Capella and I decided to include it here to add to the records of fake news put out by Russian media and pro-Russian supporters with regards to events in Ukraine, the Black Sea and Mediterranean.

To quickly recap, Russian news media outlet Izvestia claimed on 14 February 2022 that Russian navy Project 677 Lada-class SSK Sankt Peterburg had entered the Mediterranean Sea over the previous weekend “as part of large-scale exercises of the Russian Navy”. They quoted a source in the Russian defense department, stating “Together with a detachment of ships of the Northern Fleet, she will take part in manoeuvres in conditions “close to combat”[sic] “.

For this to have happened without being noticed is impossible. To have exited the Baltic Sea the SSK would have had to have transited via one of two routes between Sweden and Denmark – either via the Storebaelt bridge or the Oresund. It would also have had to have remained surfaced for the entirety of the transit. Had it done so it would have been seen either by the numerous ship enthusiasts that regularly take photographs of Russian – and other – warships; or by several webcams that operate on and in the vicinity of the Storebaelt bridge. There is no such evidence from these sources.

Many of us said the above at the time, both privately, and on Social Media. Covert Shores ran much the same story as here on the day without any satellite imagery – it was that quickly dismissed as fake news!

I requested an imagery collection from Capella Space almost immediately, and they were able to produce imagery at the next pass available, which was first thing in the morning UK time on 17 February.

The imagery provided clearly shows Sankt Peterburg still at its usual mooring position in Kronstadt, along with at least one Kilo-class SSK on the opposite side of the jetty. It’s highly likely another Kilo is tied up alongside the Lada-class.

Kronstadt has had near 100% cloud cover for well over a month making the collection of EO imagery from sources such as Sentinel impossible to use to verify the movements around the base.

This at least finalises the story as exactly what it was – a story.

More information on “Bastion-P” Matua Island deployment


  • Deployment day is now known
  • Maxar and Sentinel capture deployment
  • Identity of ships involved now known

Following on from my previous blog on the deployment of Russian K-300P “Bastion-P” mobile coastal defence missile unit to Matua Island, further satellite imagery from Sentinel and Maxar has been found showing the deployment taking place.

Moreover, the Maxar imagey has captured the actual moment the equipment exits the Pacific Fleet Project 775M Ropucha class Large landing ship – and in such detail you can clearly see the personnel filming the event.

The first thing ascertained is that the deployment took place on November 16th 2021, so the Russian MoD took two weeks to release the news. It is also now confirmed that all the support vehicles and personnel deployed at the same time rather than at an earlier date – which I suggested may have been the case in the previous blog.

If you watched the video from the Russian MoD in my last blog, at the beginning of it a Monolit-B mobile coastal surveillance radar vehicle exits the Ropucha class landing ship. The Maxar image clearly shows the Monolit-B on the beach having passed the “film crew”, and a K-300P just exited the ship and still in the breakwater.

Satellite image ©2021 Maxar Technologies.

The Ropucha can also be clearly identified as Admiral Nevelsky [055].

Imagery of the base for November 16th – the same as that from the ROLES article mentioned in my previous blog – shows little going on so it is likely that a small group of personnel were already there to set up the base, but very little else. However, it is good to get a high resolution of the image now.

Satellite image ©2021 Maxar Technologies.

Of interest is just how lucky a capture this was. Going through Sentinel-2 imagery, the weather before the 16th, up to today (December 9th), has been cloudy for the vast majority of the time. This is the drawback to normal EO imagery, and is why the SAR imagery capabilities of Capella is important to modern intelligence gathering using satellite imagery.

The Capella image from the 3rd December was collected during a period of 100% cloud cover over the island.

There is absolutely no doubt that both capabilities work in tandem with each other and will make it exceptionally difficult to hide deployments such as these in the future. Capella can provide continuous coverage of a target of interest, and should full EO imagery be required for confirmation of activity and/or actual identities of ships etc. then this can be tasked by the likes of Maxar when weather permits.

The Sentinel-2 imagery, though of low resolution, also revealed the other ships used in the deployment. These consisted of:

  • Project 141 Kashtan class tender KIL-168
  • Project 23470 salvage tug Andrey Stepanov
  • Project 19910 AGS Viktor Faleev

S-AIS data from FleetMon for Andrey Stepanov shows that she arrived at the island on the 14th November, staying for the deployment. She then left the region back to Petropavlovsk-Kamchatsky on the 17th, before returning at the end of the month and arriving on the 27th via Severo-Kurilsk at the island of Paramushir.

A quick hop back to and from Severo-Kurilsk took place over a couple of days, and she is now enroute back to Petropavlovsk-Kamchatsky as we speak.

The FleetMon Sat 15 package was worth its weight in gold for recovering this data

With this activity from Andrey Stepanov and the timing of the Russian MoD news, can one presume that the deployment is now over and just lasted two weeks?

With the Capella imagery showing very little activity at the base on 3rd December, it may well be.

But, it could also be that Andrey Stepanov has been shuttling supplies back and forth to the island – though as a tug is more likely to be there in support of a further ship such as KIL-168. With no S-AIS data available for both KIL-168 and iktor Faleev it could be they were there also. The runway is operational and also capable of taking flights for supply purposes.

For the time being, some further monitoring of the island is required.

Russian K-300P “Bastion-P” deployment to Kuril Islands


  • Russian Mod News outlet shows K-300P deployment to Matua Island
  • Capella Space imagery collection request submitted for next available pass
  • Imagery from just over 26 hours later collected and analysed

The Russian Ministry of Defence produced a small video on 2nd December 2021 of a K-300P “Bastion-P” mobile coastal defence missile unit deploying to Matau Island – part of the disputed Kuril Island chain in the Pacific.

Though only just over 1 minute and 10 seconds long, a few things can be taken from it and analysed.

The original video on the Russian MoD site can be found here or alternatively it can found on the Zvezda news online website.

The video commences with a Monolit-B mobile coastal surveillance radar vehicle (mounted on a MZKT-7930 chassis) exiting a Pacific Fleet Project 775M Ropucha class Large landing ship onto one of the beaches of the island.

This is easily identifiable in Google Earth, located at 48° 2’49.30″N 153°13’13.19″E

The complete K-300P battery – consisting of 4 launcher vehicles (TELs) and two Monolit-B radar vehicles – are seen transiting south along the beach, probably to an access track at the far end that allows vehicles to proceed up onto the mainland.

From here the battery heads east to the airfield, along the northern edge of the runway, before heading north to what was an old scientific research base located at 48° 3’59.89″N 153°15’37.12″E

Google Earth imagery from September 2019 shows that the base had been modified from the previous 2016 imagery. This was the same as the runway and was carried out during research expeditions in 2016 and 2017.

Interestingly, the battery convoy doesn’t show any support vehicles, and it is highly likely these had already arrived prior to this part of the filming. Moreover, the video shows that the research area has been fully converted into a small military base made up of trailer buildings for accommodation and operations along with a separately fenced off communications area containing at least Satcom.

On the 22nd November 2021, the Japanese Research Center for Advanced Science and Technology (ROLES) produced a report saying the research base had been recently upgraded. Imagery in the report, from Maxar, showed that the base was still in the 2019 Google Earth configuration in September 2021, but by October it was well underway to becoming the new base.

Image from ROLES report. Original imagery from Maxar Technologies Secure watch

By November 16th it appeared to be complete, including two hangers that ROLES had measured at 30 metres in length. This is sufficient to take two K-300P TEL vehicles each – though it would be a tight squeeze.

Image from ROLES report. Original imagery from Maxar Technologies Secure watch

Please head to the ROLES website to view their report.

A short sequence shows inside what is probably one of the command vehicles though nothing of note is discernible. A keypad to the right of the screen states “INTERNAL COMMUNICATION” (ВНУТРЕННЯЯ СВЯЗЬ) used for secure coms between all the vehicles of the battery.

The remainder of the video shows two of the K-300P TELs (TEL 211 and TEL 214) taking a small tour of the South-eastern part of the island before deploying into two revetments back at the airfield.

48° 2’15.12″N 153°16’32.28″E

Capella Space Imagery

Whilst the news from the Russian MoD was interesting, it didn’t actually state any dates for when this deployment took place and whether it is still in operation. It had to be after October 23rd as the Maxar imagery in the ROLES report showed the base still under construction.

The Zvezda article had quotes from the Pacific Fleet: “On this remote island in the central part of the Kuril ridge, the Pacific Fleet’s missilemen will be on a 24-hour watch to monitor the adjacent water area and straits,” and “For the operation and maintenance of equipment, the equipment of technical posts has been installed, storage facilities for equipment and materiel have been deployed, entrances to the launch sites have been equipped,”

It also stated that work has been completed on the improvement of premises intended for year-round service and residence of personnel.

I wanted to check whether this was the case and so I put in an image collection request with Capella Space on the 2nd December 2021 at 1000 GMT.

From this request they were able to create a collection task for the 3rd December at 1236 GMT when the next pass took place – just 26 and a half hours later.

From comparing the video to revetment locations, it was a simple task to find where TEL 211 and 214 positioned themselves. This can also be easily done using Google Earth despite the age of the imagery available at this time. Very little has changed here except for the base.

However, the ability to task specific collection areas for future satellite passes meant an up-to-date 50cm resolution image was available for closer inspection.

Firstly, looking at the positions taken up by the TELs in the video these are now empty, though as a mobile system this is to be expected.

TEL 211 position

TEL 214 position

Main base

The main base does still show some sort of presence, though it is impossible to ascertain whether it is actually manned or not.

Whilst vehicles are present to the northeast of the hangers this doesn’t necessarily mean that they are being used. The Russian military store vehicles at locations for ease of a quick deployment should they decide to man them.

Though not shown here, the image available to me was for the complete southeastern area of the island and therefore all the areas commonly used when a deployment takes place. Again, there is little evidence to show any activity. There are no vehicles located at any of the revetments – likewise at the various locations to the north and south that could be used for not only the K-300Ps TELs but also any of the support vehicles carrying out any other tasks.

The concrete pad area is likely for supply storage, and this does appear to be empty.

However, with an +11 hour time difference between GMT and the region, this would mean that the local time when the collection took place was 2336 and therefore everything would probably be closed down for the night and little activity would be taking place anyway.

The fact that it was 2336, and therefore night time, highlights how useful the Capella SAR capability is. I was able to put in a request, and they were able to set up an image collection at the next available pass – despite it being dark. It could have been days before the next EO pass could be available from other providers.

This region is one of those that I regularly check so it will be interesting to keep an eye on it through imagery from Capella from time to time.

K-300P “Bastion-P” basic information

  • Anti-surface mobile coastal defence system
  • Consists of four TELs, each with two P-800 Oniks missiles, two Monolit-B mobile coastal surveillance radar vehicles, up to four loader vehicles, command and support vehicles.
  • Monolit-B has an approx range of 250 km in active mode, 240 km in passive mode
  • Can track up to 50 targets in passive mode, 30 in active
  • P-800: 6.9m in length; can cruise up to 2700 km/h at altitude; range between 120 km and 300 km depending on flight profile used; High cruise 46,000 ft, Low cruise 22 ft; 200 kg warhead
  • Command vehicle crew: 5
  • TEL crew: 3

Further information on Matua Island

The island itself is steeped in history, particularly for World War 2. All the trenches shown in the imagery are from that era. There’s a whole network of Japanese tunnels and bunkers located throughout the island; and reports that it was a potential test area for their nuclear weapons program that was being carried out at the end of the war.

A Zvezda video from 2017 on YouTube is well worth a watch – although it is obviously in Russian.

There’s also plenty of blogs and articles online that describe some of the research trips that have taken place there. To list a few:

https://zizuhotel.ru/en/bangkok/sekrety-matua-chto-skryvayut-nedra-kurilskogo-ostrova-yaponskuyu-krepost/