Miyako Island is an island of Okinawa Prefecture located about 300 km southwest of Okinawa Main Island. It has an area of approximately 158.5 km² and a population of about 55,000.
The entire island is formed of limestone derived from uplifted coral reefs, and this geological characteristic has created the diverse cave environments unique to Miyako Island.
Because limestone is easily dissolved by rainwater and groundwater, limestone caves form over long periods of time. In coastal areas, caves are also formed by dissolution caused by seawater. As a result, Miyako Island has many caves, both on land and underwater. Some land caves are known as tourist sites, and many underwater caves attract attention from divers in Japan and abroad.
In this project, among these various caves, we focus on the “subterranean lakes” located in the deeper parts of land caves and explore their underwater environments.
The geological structure of Miyako Island can be broadly divided into the following three layers:
Because the highly permeable Ryukyu Limestone overlies the relatively impermeable Shimajiri Layer, groundwater tends to accumulate near the boundary between them, and subterranean rivers and subterranean lakes are thought to form there.
These strata are not distributed uniformly in horizontal layers. In some places, the Ryukyu Limestone is exposed at the surface, and in other places the thickness of the layers changes significantly.
In addition, faults running roughly north-northwest to south-southeast are developed on Miyako Island, and groundwater is believed to flow along these faults, appearing as springs and subterranean lakes in the deeper parts of caves.
These geological and groundwater conditions made possible the world’s first underground dam on Miyako Island, which forms the basis of the island’s water resource management.
References on geology and groundwater consulted for this research are listed at the end of this page.
※ Imgya Marine Garden, Miyako Island
※ Fukuzato Underground Dam
※ Toriike, which connects to the sea underwater
It is widely known that Miyako Island and the neighboring islands of Shimoji and Irabu contain many land caves and underwater caves.
Even in recreational diving, Miyako Island is so well known for cave diving that sites such as “Maou no Kyuden” and “Toriike” attract many divers from Japan and abroad.
At the same time, many subterranean lakes in the deeper parts of land caves have a history of being used as domestic water sources, and traces of daily life remain inside the caves.
For this reason, many of these caves and subterranean lakes have been designated as cultural properties or historic sites, and underwater exploration has generally not been carried out.
In that context, Professor Yoshihisa Fujita of Okinawa Prefectural University of Arts, with whom I had previously conducted cave research in other regions, consulted me about exploring the deeper underwater sections of subterranean lakes on Miyako Island that he had not been able to reach because of technical diving limitations.
Exploring uncharted areas that require advanced diving techniques is precisely the kind of challenge I should undertake as an underwater explorer, and after discussion with Professor Fujita, we decided to carry out this research.
At that time, the idea was that the work would proceed “if research funding could be obtained in the following fiscal year.” However, in discussing this, Professor Fujita and I realized that we shared the same concerns about that issue and about the broader way basic research is often conducted in Japan.
※ Nudokubi Abu on Irabu Island. Traces remain of its use by Japanese troops during the war.
During the planning of this research, we became aware of two major issues in conventional research methods.
Researchers are highly skilled at evaluating and studying the academic significance of information obtained from research sites. However, in environments such as underwater caves, advanced diving techniques and careful safety management are essential, and this greatly limits both the number of researchers who can dive and the range that can be explored.
Underwater explorers, on the other hand, possess high-level diving skills and the ability to carry out field exploration, but there are limits to how far they can independently evaluate and communicate the academic significance of what they find.
Therefore, we believe that collaboration between researchers and underwater explorers could produce new research results by combining the strengths of both.
In general, many researchers in Japan have conducted research using funds allocated by their institutions and external research funding such as KAKENHI and private research grants.
Institutional research funds are often limited, while external grants have low acceptance rates and require a long time from application to decision. When applications are unsuccessful, repeated reapplications may be necessary, which has also been pointed out as a waste of researchers’ human resources.
This is especially true in biodiversity studies, where it is impossible to know in advance what organisms may be found in the field. Ideally, joint fieldwork involving taxonomists from multiple disciplines would be desirable, but because of limitations in budget and personnel, such fieldwork is often difficult to carry out.
These two issues can ultimately be summarized as the inability to secure flexible funding for sufficient work in the initial stage.
For that reason, in this project we explored a new framework: to carry out initial exploration efficiently, make the research value visible at an early stage, and then connect that value to funding for the next phase of research.
More specifically, we are considering crowdfunding and small-scale sponsorship from individual supporters and companies by offering, for example:
If this system works, it could make it possible to extend exploration to places that were previously out of reach for budgetary reasons, and it may accelerate future research outcomes.
For the March 2026 exploration, we selected the following three caves.
Miyako Island has many other springs and subterranean lakes, and we intend to extend the scope to them in the future.
The subterranean lakes in all of these caves are thought to be anchialine environments*, and are expected to contain previously undescribed species and other unusual organisms.
*Anchialine environments are defined as tidally influenced underground estuaries in the fissures and cavities of karst and volcanic terrain, extending inland as far as seawater penetrates (Bishop et al. 2015, Journal of Crustacean Biology 35(4): 511-514).
Pottery and bones believed to be human have already been found underwater in Kinsukya-ga, and further archaeological and biological discoveries may be expected in the unexplored sections deeper inside the cave.
Pikyazu is a doline that opens through to the sea. According to a report published in January 2017 in the Bulletin of the National Institute of Technology, Sasebo College, No. 53, the research team was able to proceed to a point about 20 meters from the water surface and about 11 meters in depth, and confirmed water flow from deeper inside, but had to stop there because of silt accumulation.
In this project, we aim to confirm whether it is possible to reach beyond that point. Regardless of whether that proves possible, we also plan to conduct biological research focusing on the mixing zone between the lower-salinity flow coming from deeper inside and the more saline seawater.
※ Near the water surface of Pikyazu
After discussions with Professor Fujita and members of JKUEP (Japan Karst & Underwater Exploration Project), the exploration period was set for March 4 to 9, 2026.
To address the issues described above, Atsushi Hayashida, who can carry out surveying in land caves, Yudai Uchida, a cameraman capable of underwater and topside imaging, and archaeology researchers were also scheduled to take part in this research.
The two of them are also my technical diving students, and I am glad that we can now work together in this way, beyond the teacher–student relationship.
As described below, the scope of work was defined in advance. Among these items, underwater cave surveying and 3D mapping are especially important. Although they provide fundamental data for academic research, there are still few specialists in Japan with these skills, and we believe the data obtained in this research will make a significant contribution to future studies.
Ahead of the main work, in November 2025 I carried out a preliminary field exploration with underwater photographer Jun Shimizu, after filing the necessary notifications with Okinawa Prefecture and Miyakojima City.
This was done to secure transport routes for diving equipment and to confirm the environment in advance. It may sound excessive, but I believe this kind of preparation is extremely important for efficient and safe exploration.
※ Beautiful water surface of Kinsukya-ga
※ Inside the cave of Amaga of Tomori, and its water surface
On that day, NHK staff who were interested in the exploration came along with Jun Shimizu and me. Although the weather was poor, we inspected the sites and considered routes and logistics.
The water surface of Kinsukya-ga was remarkably beautiful and made me want to dive in right away.
Of course, its importance is not only aesthetic. Pottery has been confirmed on land, and in the water, pottery as well as bones believed to be human have been found, making it a cave of great academic value.
The water surface of Kinsukya-ga rose and fell with the tides, suggesting, as prior information indicated, that the spring is connected to the sea.
At Amaga of Tomori in Gusukube-Cho, Professor Fujita discovered larvae of marine organisms even though the site is inland.
The presence of larvae of marine organisms strongly suggests that this site too is connected to the sea underwater. Further biological research here may contribute to clarifying how organisms inhabiting such anchialine environments disperse.
In addition, the steps inside Amaga of Tomori, which were installed to draw water, reminded us that this water was once used in daily life.
We also received information about the sites from staff of the Miyakojima City Board of Education. I would like to record my sincere thanks here.
※ From the NHK Okinawa broadcast of November 13, 2025
Related coverage (November 2025 preliminary exploration)
Following the preliminary exploration in November 2025, the team assembled on Miyako Island on March 4, 2026: Atsushi Hayashida, Yudai Uchida, and I for the diving, together with Professor Yoshihisa Fujita (Okinawa Prefectural University of Arts) and archaeologist Seita Gushiken (Okinawa Prefectural Board of Education) on the research side. With permission from Okinawa Prefecture and Miyakojima City, we carried out underwater exploration at three sites: Kinsukya-ga, Pikyazu, and Amaga of Tomori.
On the day before the exploration, the whole team met to go over research methods and prepare the equipment.
We dived Kinsukya-ga on March 5 and 8, and Pikyazu and Amaga of Tomori on March 7.
※ All members of the March 2026 exploration
※ Underwater in Kinsukya-ga
※ Okinawan ceramic ware
In November I wrote that the water surface of Kinsukya-ga made me want to dive there immediately. On March 5, we finally did.
When I tasted the water on entering, it was saltier than I had expected, which hinted at a connection to the sea.
The water was clear, with visibility of 10 to 12 meters, and there was not much silt on the bottom.
The main passage is no deeper than 5 meters, and the deepest point is about 10 meters. As we laid line (a thin rope that divers lay through a cave to secure their route in and out) through the whole cave, including the narrow passages, and moved deeper in, bones and vessels kept appearing within reach of our lights.
Bones believed to be human, and vessels, lie scattered from near the entrance to the far end of the passage, in quantities too great to describe in full.
Gushiken examined them for us and gave his opinion that some of these vessels are Miyako-type pottery. Underwater we also observed objects that appear to be ornaments.
Why are bones and vessels scattered all the way to the far end of the passage?
Miyako is known to have had a tradition of cave burial (fuso, or exposure burial), and it is not unusual to find bones in a cave with pottery placed beside them.
Was Kinsukya-ga, then, once a burial cave? If it was, two things do not fit.
The first is that the bones are underwater at all. Burial grounds do sometimes become submerged, but that happens when the water level rises for some reason.
Miyako-type pottery was made from the medieval to the early modern period (roughly the 13th to 18th centuries), and it is unlikely that the water level rose rapidly after that.
The second is that the bones are scattered to the far end of the passage and even to the far end of a side passage. In exposure burial, remains are not usually found scattered throughout a cave.
How, then, did they get here? There are two possibilities: someone placed them there directly, or they were carried in by water.
※ Bone-like objects observed inside the cave
※ Removing tanks in the air dome to climb onto dry land
We began with the first possibility. If there were another entrance opening to the outside partway along the passage, the chance that the remains were placed directly in the far end of the cave would increase.
There was a clue. Underwater in the cave there is a place where an air dome allows you to climb onto dry land, and bones can be seen on that land as well. Insects were also observed there, and the presence of large insects strongly suggests a connection to the outside.
On March 8, Hayashida removed his tanks, climbed onto the land, and checked. At least at present, there is no passage large enough for a person to pass through. That made it less likely that they were placed there directly.
The far end of the main underwater passage appeared to be blocked by a collapse. We threaded our way between the fallen rock as far as we could, but in the end the gap was no longer large enough for a person to pass.
Space probably continues beyond the collapse, but for now there is no way to get there. We also confirmed two side passages: one that drops to about 10 meters and silts up easily, and a narrow one that branches off to the side.
Vessels and bones were observed at the far end of the deeper side passage as well. Even if, by some chance, this side passage was not submerged in the past, it is hard to imagine that anyone came this far into a side passage to lay remains to rest.
The total length of line we laid during this exploration, including the side passages, is roughly 120 meters.
Kinsukya-ga has been dived several times in the past, but apart from the work of Professor Fujita, who joined this exploration, no academic research on its underwater section has been reported.
There is no other entrance, and it is hard to believe they were placed there directly. What remains is the possibility that they were carried in by water.
The water in Kinsukya-ga is not fast-flowing groundwater like an underground river; it is almost still.
We felt almost no current while diving, and the everyday movement of the water is limited to a slight flow that follows the tides.
If, in such an environment, the remains were carried deeper in, it was not by everyday movement: some special event must have occurred.
Gushiken, who accompanied us, raised the possibility that the Great Meiwa Tsunami of 1771 carried remains from near the entrance deeper into the cave, and further, that remains may have been washed in from burial grounds outside the cave.
If so, this cave would be not only a burial site but also a record of the tsunami. We hope that, as we gather further information, specialists will be able to clarify this.
Our role is to bring information back from underwater and hand it to researchers as material they can judge. This record is the first full set of that material.
The saltiness I had tasted showed up directly in the numbers.
Starting at 11:40 on March 8, we sampled water at eight points from the surface down to 10 meters. Salinity was 7.62 PPT at the surface, 9.06 at 1 meter, 11.61 at 3 meters, 14.10 at 7.5 meters, and 15.46 PPT at 10 meters. Rather than a boundary at some depth, the water simply becomes steadily saltier with depth, and we found no sign of stratification. This matches what we saw with our own eyes. We felt no current while diving; pH was 7.08 to 7.24, and the water temperature was 23 to 24°C.
Some of the small crustaceans that live in anchialine caves were observed underwater, but not in great numbers.
These are observations from a single day in March. The values are likely to differ with the time of measurement, depending on rainfall, season, and the tides.
※ Part of the measurement data from this exploration
※ The water surface of Pikyazu
※ Following the flow of water into the cave
Pikyazu is a feature in the sea resembling a collapse doline. We swam out from the beach with underwater scooters and approached it from the seaward side.
Our original plan was to focus the biological research on the zone where flow from deeper inside meets seawater. Once in the water, however, we found the water being drawn strongly inward toward the far end. The marine influence was so strong that this was not an anchialine environment in the usual sense.
With Hayashida leading, we began the dive and reached the point at about 11 meters' depth that the 2017 report by Hiroki Manabe and colleagues of the National Institute of Technology, Sasebo College gave as their farthest point. It looked possible to go further, but from there the passage narrows, and the inward suction becomes even stronger.
If all three of us went in at once, anyone who became stuck might not be able to get back out, so Uchida and Hayashida waited while I went ahead to check whether the passage continued.
In a narrow passage, going in head-first without knowing whether there is room to turn around is extremely dangerous.
Feet first, I worked my body slowly into the gap. At around 20 meters' depth, I found water being drawn with great force into a gap between rock above and below, about 20 to 30 centimeters high and about 1 meter wide. It is not large enough for a person to pass, and this became our farthest point.
The 2017 report describes flow coming from deeper inside; on this dive, the flow was being drawn inward.
The part a person can pass through ends here, but the water keeps being drawn in beyond it. The cave interior is a fully marine environment, and the organisms are the same as those in the surrounding sea, so it did not seem worth pursuing further as a subject of this research.
This is one settled record about this place: the next person to ask a question about Pikyazu can start from here.
In that sense, I think we have done our part.
※ Team members waiting while Isaji checks the far end
※ Isaji emerging from the narrow restriction
※ The water surface of Amaga of Tomori seen from land
The subterranean lake of Amaga of Tomori is a well at the bottom of a vertical shaft, and from land it was not certain that the spring continued far enough to be dived.
On March 7, the day of the dive, Hayashida free-dived to check the back of the spring and confirmed that a cave continued beyond it, so we decided to dive.
We were told that, as far as the records of the boards of education of Okinawa Prefecture and Miyakojima City show, this is the first permit for an alteration of the existing state involving diving. Underwater there was no trace of previous diving other than the line we had laid, so we consider this to have been the first dive in Amaga of Tomori.
From the back of the spring the bottom slopes down to about 3 meters, and from there it is level.
Silt has settled on the bottom; rather than level ground, it is level because of the silt that has accumulated there.
Following the left wall for about 15 meters, we came to a chamber about 10 meters wide, with fine stalactites. It is a splendid underwater cave.
Straight on from there, the passage dead-ends after about 15 meters. This dead end was confirmed by Hayashida and comes from his report; I was the last of the three and waited short of it because of the silt-out.
My feeling is that, as reported, it does not extend much further, but we will check again next time.
※ Underwater in Amaga of Tomori
※ The floor of Amaga of Tomori; crustaceans are visible
This cave silts out very easily.
In addition to the thick silt on the bottom, percolation (sediment dislodged from the ceiling by exhaled bubbles) is severe, and if you stay in one place the visibility is gone in no time.
On the way out, of course, visibility was zero.
Even when it is not silted out, visibility is only about 5 meters, so this time we checked along the left wall. The right wall is a task for next time.
Organisms, on the other hand, were plentiful: we observed a great many of the crustaceans that live in anchialine environments.
This suggests that, because Amaga of Tomori is a vertical shaft, a large amount of organic matter enters the water, which means plenty of food.
The water seemed fresher than Kinsukya-ga; this is an anchialine environment, but the connection to the sea appears to be weak.
Next time, if time allows, we would also like to observe how the water level changes with the tides.
In this exploration, we completed dives at all three sites without incident and gained a picture of the different cave environments that exist within a single island.
At Kinsukya-ga and Amaga of Tomori, I think we have also seen what our next tasks should be.
Amaga of Tomori in particular appears to have been dived by no one but us so far, so there is no underwater information about it anywhere. We want to give priority to gathering basic information there and reporting it to the boards of education and other authorities.
Miyako Island and the neighboring islands have many other caves, and some places are likely to have subterranean lakes that can be dived.
We hope to widen the scope of our activities and explore them as well.
Related coverage (March 2026 exploration)
Following the March exploration, we returned to Miyako Island in September for further underwater exploration. This time, permission for alteration of the existing state was granted in my name by the Okinawa Prefectural Board of Education (Amaga of Tomori in Gusukube-Cho) and the Miyakojima City Board of Education (Kinsukya-ga and Muika-ga). We widened the scope by one site, adding Muika-ga (盛加がー), a Miyakojima City-designated Historic Site.
The participants were Yudai Uchida, Sanae Hayashi, Dan Mibayashi, who provided support such as carrying equipment, and myself. There was light, intermittent rain. At every site, our records were made by photography alone, and we did not touch any artifacts.
※ Muika-ga: the pool right at the bottom of the stone steps. Vessel fragments can be seen both on land and underwater
Muika-ga, in Hirara, is a cave well of the kind known as an uri-ga. Miyakojima City’s description of its cultural properties notes that, before piped water became widespread, such uri-ga served as sources of domestic water; that Muika-ga is the largest in the Hirara area, with 103 stone steps; and that there is a small shell-midden layer inside the cave, with celadon and earthenware fragments to be found in the surrounding area**.
Right at the bottom of the stone steps is a pool. About 5 meters further in, across a stretch of land, there is another, smaller pool; the two are separated only by rock, but their water is connected. Both pools have a water temperature of about 25°C and are shallow, about 50 centimeters to 1 meter deep, and the water is very clear. For the first part of the exploration, we observed underwater by eye with only masks and fins, without tanks. We were looking for artifacts and organisms, and for whether a passage continued deeper underwater.
Looking through a gap in the rocks of the nearer pool, at the point where the two pools connect, we could see a small space below. Thinking that a passage might be found from the other side, that is, from the inner pool, we checked from there and found a narrow passage descending about 2 meters at an angle of about 45 degrees. It was just about the size of a body.
Simply putting our faces in the water was not enough to see to the end of the passage, so we had a tank brought over to check it.
Because the passage was narrow, I did not wear the tank: leaving it on the bottom, I ran out the 2-meter long hose (the regulator hose) and put only my head into the passage to look.
Beyond the passage I found a small space, by eye about 80 centimeters wide, 80 centimeters deep, and 60 centimeters high. The depth there is about 2 meters.
To check whether a passage continued from this space, we attached an extension pole to a 360-degree camera, put it in, and filmed. There is no need to put one’s whole body in until a passage is known to be there. Putting your whole body in would stir up the water and leave you unsure which way to go, so it is more reliable to get a rough sense of direction first.
As a result, we found that there is no passage beyond this space: the subterranean lake of Muika-ga does not continue further in at a size a person can enter. This could be confirmed only by diving, and it stands as a record of this place.
**Miyakojima City, 「【市指定：史跡】盛加がー｜宮古島市の文化財」 (Miyakojima City website, in Japanese)
※ Leaving the tank on the bottom instead of wearing it, and checking beyond the gap in the rock on a long hose
※ Beyond the passage, filmed with a 360-degree camera: only a small space, with no passage continuing
※ Vessel fragments in the nearer pool
We felt no movement of the water at all, yet the water level dropped by several tens of centimeters between the start and the end of our exploration. It spanned the time of low tide, and the level seemed to follow the change in sea level with a slight delay.
Although the water is largely fresh, taking together the change in water level within a single day, the position of the cave, the approximate height of the water surface, and its location on Miyako Island, we judged that Muika-ga is also an anchialine environment.
Right at the bottom of the stone steps, vessel fragments were seen both on land and in the water of the nearer pool. There were many fragments underwater in the nearer pool, and we had to take care not to break or move them as we moved around.
In addition to ordinary photographs, we photographed the fragments with a scale.
No bone-like remains were seen.
In the inner pool, on the other hand, there were no fragments underwater. In this respect it differs from Kinsukya-ga, where artifacts were observed all the way to the far end of the subterranean lake. Given its position, we believe this is because no external force such as a tsunami, capable of carrying fragments further in, acted on Muika-ga.
We saw no organisms underwater by eye, but had the impression that some might appear if traps were set.
When we entered Amaga of Tomori, the water level was higher than before.
The water was very full this time, and we strongly felt the change in water level since March. As prior information had indicated, this too is an anchialine environment. The water temperature was 25°C.
The water was far more turbid than last time, and from the start of the dive we could see almost nothing.
Near the bottom, suspended silt left almost no visibility, and we had to give up the survey we had hoped to carry out this time.
When we added line and went to check the upper part of the water, however, we could see about 2 meters, so we decided to explore the upper part of the chamber we had found in March.
The upper part of the chamber forms a small air dome along the right wall, about 50 centimeters high, where you can put your head out of the water, and many stalactites were observed both in the air dome and in the chamber.
Although it is called an “air” dome, the air in such a small space is not necessarily breathable, so I checked it without taking the regulator out of my mouth.
The right wall, which we had left as a task in March, was also checked to some extent during this exploration.
※ Some visibility remained in the upper part of the chamber
※ Well-developed stalactites in the chamber
Partway through the chamber, part of the line laid in March was buried in silt.
Enough silt to bury it has newly accumulated over the past six months.
Given how much silt has accumulated, the passage may once have continued further in.
At the same time, as the years go by, the areas we can enter now may also become filled in, and I felt once again how important it is to record the present state.
At Kinsukya-ga, this time we photographed each location and the artifacts.
In addition to ordinary photographs, we photographed the artifacts with a measuring tape so that their size can be read from the images. Visibility was the same as in March, and the water temperature was 25°C, warmer than I had expected.
About 10 meters in from the water surface of the subterranean lake, beyond the slope that continues from the lake, we confirmed one complete vessel.
Gushiken has given his opinion that this vessel is Okinawan ceramic ware. This means that, underwater in Kinsukya-ga, there is Okinawan ceramic ware as well as Miyako-type pottery.
In addition, beyond a narrow passage leading from the subterranean lake in the opposite direction from the main passage, we confirmed a vessel of unknown age.
We also found what appear to be skulls underwater, at two places: almost at the far end of the main passage, and at the far end of the narrow passage. Both are places where, we felt, things would collect if washed in by a large volume of water.
※ Complete Okinawan ceramic vessel
※ Stalactites in Kinsukya-ga, reminiscent of the caves of Minamidaito Island
We also observed stalactites in the cave. Their form, color, and size are all very impressive, and they reminded me of the stalactites I have seen in the caves of Minamidaito Island.
This time we felt a current that we had not felt in March: a slight flow drawing water into the side passage that drops to about 10 meters and silts up easily. It is still within the range of a slight flow, but it was stronger than last time; the variation over time that we anticipated in the March section has shown up as a current.
From next year onward, we plan to produce a survey map and plot the artifacts on it.
We have compiled the results of this exploration into a report and submitted it to the Miyakojima City Board of Education.
We would be very glad if our exploration proves useful for future research and activities.
*Affiliations may not be up to date.
Copyright © 2024 Japan Karst & Underwater Exploration Project