At 16:27 on 28 July of Reiwa 8 (2026), a magnitude 7.1 earthquake struck Kumamoto. As of this writing, a full three days have passed. On 29 July — the day after the quake — HotelBank published two pieces: one tallying the guest-room stock and epicentre distance bands across the 21 municipalities covered by the Disaster Relief Act, and one revisiting the demand asymmetry of 2016. This article is the follow-up, and it presents what has changed in the 72 hours since, as a comparison between the deltas in official announcements and re-aggregated values for the same indicators.

It also confronts head-on two questions the previous two articles deliberately left alone — “which properties have officially announced closure or suspension of operations?” and “does the disappearance of inventory in lodging data mean a property has closed?” To state the conclusion first: the two do not match at all. The errors run in both directions, and reading post-disaster supply-demand indicators without understanding that structure will badly misrepresent what is actually happening in the market.

1. What Changed Since the Last Report — How Official Announcements Were Updated in 72 Hours

First, we place side by side MLIT’s “Report No. 5” (as of 14:00 on 29 July), on which the previous two articles relied, and the latest “Report No. 12” (as of 13:30 on 31 July). The same bulletin format has now been updated 12 times, making it possible to read off which items improved and which worsened.

Source: Compiled by the HotelBank Editorial Team from MLIT, “On the Damage Situation Caused by the Reiwa 8 Kumamoto Earthquake,” Report No. 5 (as of 14:00, 29 July 2026) and Report No. 12 (as of 13:30, 31 July 2026)

From a lodging-demand standpoint, the single biggest change is the resumption of Kyushu Shinkansen service between Hakata and Kumamoto. In the previous article we wrote that the Kyushu Shinkansen was suspended along its entire length, effectively making all of Kyushu “hard to reach” for inbound travellers using rail passes. With that section reopening on 31 July after three days, the rail corridor from Fukuoka to Kumamoto has at least been restored. JR Kyushu notes, however, that it is running a reduced number of services, and through-services with the Sanyo Shinkansen remain suspended. Between Kumamoto and Kagoshima-Chuo, more than 200 locations of fallen or damaged noise barriers, two rail fractures and one displaced bridge girder have been confirmed, and the company states that “detailed damage is still under investigation.” A restoration plan for the Kagoshima Main Line between Kumamoto and Yatsushiro is also still being drawn up.

Seismic intensity determinations have also been updated. The table in the previous article listed upper 6 intensity for five municipalities — Kumamoto City, Yatsushiro City, Uto City, Misato Town and Mashiki Town — but Report No. 12 adds Kashima Town, bringing the total to six. Kosa Town has also been added at upper 5. Kashima Town, which the previous article classified as lower 6, is now published as upper 6. The fact that such updates continue as provisional values are refined is a characteristic worth bearing in mind when publishing anything grounded in seismic intensity.

The scale of casualties and evacuation has changed as well. At the time of the previous article (07:00 on 29 July, Cabinet Office), the number of evacuees was 9,186. At Kumamoto Prefecture’s disaster response headquarters meeting on 31 July, it was announced that 35 people have died, including one case where the connection to the disaster is still under investigation, and that just over 9,100 people remain evacuated. Water outages affected approximately 79,100 households in the prefecture as of 07:30 on 31 July, of which Yatsushiro City accounts for roughly 25,300. As the previous article noted, lodging capacity does not exist on an intact building alone — it is inseparable from restored water supply. Over these three days, the number of households without water has shrunk only from about 84,000 to about 79,100, a decline of roughly 6%.

2. By Epicentre Distance Band — The Inventory Signal Decays Monotonically from 97.3% to 0.0%

From here we turn to this article’s own aggregation. The previous article showed peacetime stock — how many rooms accumulate in each distance band from the epicentre. Here we use the same distance-band axis to re-aggregate how observation of sales inventory changed after the quake. Our earlier tally of how much room stock sits in each distance band across the 21 municipalities covered by the Disaster Relief Act found 13,269 rooms within 30 km of the epicentre and 17,708 rooms across 378 properties in those 21 municipalities.

The population is limited to properties in Kumamoto Prefecture for which sales inventory for August 2026 check-ins was observable on at least 10 days before the quake (before 16:27 on 28 July). Including thinly observed properties in the population would cause mere observation noise to be misread as “inventory loss.” Under this condition, 299 properties and 17,995 rooms remain. One property had a legacy record under a former name at the same coordinates; this was excluded as a duplicate.

Each property was sorted into the following four categories based on observations before and after the quake. This classification is strictly a data-level inventory signal, not a determination of whether a property is operating. Section 4 demonstrates why.

Source: Compiled by the HotelBank Editorial Team from MetroEngines Research (N=299 properties / 17,995 rooms, aggregated 31 July 2026)

Splitting these 36 properties and 3,218 rooms by distance from the epicentre produces a clean monotonic decay.

Source: Compiled by the HotelBank Editorial Team from MetroEngines Research (N=299 properties; epicentre 32.6°N, 130.7°E as published by JMA; aggregated 31 July 2026)

Source: Compiled by the HotelBank Editorial Team from MetroEngines Research

In the 0–10 km band, 182 of the 187 rooms across six covered properties (97.3%) lost their inventory signal. This band contains only six properties, however, so movement at a single property swings the rate considerably and the figure should be discounted accordingly. The 10–20 km band shows 712 of 1,293 rooms across 15 properties (55.1%); the 20–30 km band 2,178 of 9,966 rooms across 101 properties (21.9%); the 30–50 km band falls to 3.4%, and beyond 50 km to 0.0%.

What matters here is that while this decay corresponds closely to the seismic intensity distribution, the absolute volume of rooms increases with distance. As the previous article showed, there are almost no rooms to serve as capacity in the immediate vicinity of the epicentre; the lodging cluster sits about 22 km to the north, in central Kumamoto City. Of the 3,218 rooms that lost inventory, therefore, the largest block by room count is the 2,178 rooms in the 20–30 km band — that is, central Kumamoto City. By rate, the area nearest the epicentre is the most severely affected; by volume, the prefectural capital takes the largest hit. This duality is easy to conflate when relief coordinators are judging “where has capacity thinned.”

The geographic distribution is worth confirming on a map. The concentric circles mark 10, 20, 30 and 50 km from the epicentre. Blue circles are properties that have officially announced closure or suspension of reservations (Section 3); grey circles are properties confirmed to be operating normally per official announcements (Section 4). Circle size represents room count.

Source: Compiled by the HotelBank Editorial Team from MetroEngines Research (property locations and room counts), JMA (epicentre) and official announcements from each property

3. Properties That Officially Announced Closure or Suspension — 6 Properties, 830 Rooms

This is the core of the article. We name properties only where closure, suspension of operations or suspension of new reservations is explicitly stated on the property’s own official site or in an official announcement from its operating company. We do not treat the absence of visible inventory on booking sites, or a mention in news coverage alone, as grounds for declaring a property “closed.” The next section explains why.

For each listing, we also confirmed that the closure notice was issued after the 28 July earthquake, ensuring these are not pre-existing closures unrelated to the quake.

Source: Compiled by the HotelBank Editorial Team from content posted on the official sites of each property and operating company (verified 31 July 2026). Room counts from MetroEngines Research; distance from the epicentre is the great-circle distance from the JMA-published epicentre (32.6°N, 130.7°E)

The figure of six properties and 830 rooms is the range for which this article was able to verify official announcements. In reality, it is highly likely that more properties are closed. Smaller ryokan and minshuku in particular may not have their own websites, or updates may have stopped because of telephone and water outages — meaning that “an announcement” was never possible in the first place. The gap between the 36 properties flagged by the inventory signal in Section 2 and the six that could be officially verified reflects the limits of information flow itself in the immediate aftermath.

Looking at the relationship with distance, three of the four properties that officially announced closure are in Yatsushiro City and one is in Hikawa Town — all within 14 km of the epicentre. The two properties in Kumamoto City (Nishi and Chuo wards), by contrast, have taken the form of “suspending reservations” or “suspending in-house facilities and changing operations” rather than closing. The former represents a physical stoppage from building and equipment damage plus severed lifelines; the latter reads as a decision to restrict new sales while operations themselves remain possible, given constraints on water, food supply and housekeeping capacity. The very shape of a suspension differs by distance band.

4. “Inventory Disappeared” Does Not Equal “Closed” — The Errors Run Both Ways

This is the point this article most wants to stress. The “36 properties that lost their inventory signal” in Section 2 and the “six properties that officially announced closure” in Section 3 barely overlap. And the reason they do not overlap runs in more than one direction.

Error direction 1: inventory disappears but the property has not closed. Among the properties classified as “B: listing continues, zero rooms remaining” in Section 2, we checked the official announcements of the operators of the top three by room count. All three are operating normally.

Source: Compiled by the HotelBank Editorial Team from content posted on each operator’s official site (verified 31 July 2026) and MetroEngines Research (inventory observation, including a re-retrieval of listing status on the evening of 31 July)

In other words, the 651 rooms that appear to have “disappeared” in the data are in fact operating. So what is happening? The most plausible explanations are several: (1) rooms are effectively filled by long-staying recovery and relief personnel, leaving no allocation for general sale; (2) constraints on water, food supply and housekeeping capacity mean properties are prioritising existing reservations and temporarily throttling new sales; and (3) operators have deliberately withdrawn online sales allocations and switched to handling direct enquiries. Publicly available online information alone cannot determine which of these is the primary cause. When Kumamoto Prefecture’s total guest nights turned positive year on year in June following the 2016 earthquake, lodging demand from recovery personnel was part of the background — and even if the same phenomenon is occurring now, it cannot be read as “a recovery in leisure demand.”

Part of this reading was borne out within the same day. On the evening of 31 July we re-checked the listing status of the three properties: Via Inn Prime Kumamoto had resumed selling across every August date, so a signal that had read “zero rooms remaining” until early morning cleared within half a day. Washington R&B Hotel Kumamoto Shimotori and Toyoko Inn Kumamoto-jo Toricho-suji, meanwhile, had moved from listings with zero remaining rooms to withdrawing their August sales plans entirely. Operating while manipulating only the sales allocation — this progression shows that pattern (3) is in fact a major dynamic. The inventory signal is not a mapping of operating status; it is point-in-time information that reflects a property’s sales operations on a scale of hours.

Error direction 2: a property has officially suspended, yet inventory persists in the data. Of the six properties in Section 3, Mitsui Garden Hotel Kumamoto officially announced on 28 July that it had suspended reservations through 30 September — yet sales inventory for August check-ins at that property was still observed on 11 days after the quake (31 days before it). THE BLOSSOM KUMAMOTO likewise showed seven days of post-quake inventory. HOTEL AZ Kumamoto Yatsushiro Miyahara announced a temporary closure on 28 July, yet the 31 July observation still recorded inventory on eight days of sales rows. This is because reservation suspensions are reflected at different times on different sales channels, and because a property announcing a suspension does not cause the listing information itself to disappear immediately. On the evening re-check of 31 July, Mitsui Garden Hotel Kumamoto’s August listing had disappeared and THE BLOSSOM KUMAMOTO’s remaining rooms had gone to zero on all dates — roughly three days from announcement to data reflection.

5. How Large Is the Supply Loss? — Upper and Lower Bounds by Room Count

With the previous section in mind, there is still a need to grasp the order of magnitude of how much supply has been lost. This article presents two figures of differing certainty as an upper and a lower bound.

Source: Compiled by the HotelBank Editorial Team from MetroEngines Research and official announcements from each property (N=299 properties / 17,995 rooms, aggregated 31 July 2026)

The reality lies between the lower and upper bounds. The six properties and 830 rooms verified officially in this article represent supply that is certainly lost, while the 3,218-room upper bound certainly includes the 651 operating rooms identified in the previous section. Mechanically subtracting those 651 rooms gives 2,567 rooms (14.3% of the population), but similar false positives must also be present among properties we did not check — so it is reasonable to view the actual supply loss as somewhere between 830 and 2,567 rooms, on the order of 1,000–2,000 rooms.

It is worth setting this against the previous article’s figures. That analysis counted 378 properties and 17,708 rooms of stock across the 21 municipalities covered by the Disaster Relief Act, and 13,269 rooms within 30 km of the epicentre. If 1,000–2,000 rooms are in fact unusable, that corresponds to roughly 8–15% of the stock within 30 km. In the previous article’s “distance band × mobilisation rate” capacity calculation, meeting just over 10% of evacuees within the 30 km radius would require mobilising 5% of rooms — which means 8–15% has become unusable before that 5% can even be secured. Capacity planning needs to be reset to start not from total stock but from “rooms usable right now.”

6. The Apparent Rise in Demand Indicators Is the Flip Side of Inventory Disappearing

This is the most operationally dangerous point of all. In a disaster-hit area after an event, estimated occupancy (OCC) rises. But not because demand increased. When inventory is withdrawn, rooms whose remaining count has gone to zero are recorded as “sold.” Our aggregation measures this phenomenon directly.

First, we place Kumamoto Prefecture’s estimated OCC alongside the average of Kyushu’s other six prefectures (Fukuoka, Saga, Nagasaki, Oita, Miyazaki and Kagoshima), by check-in date. To cancel out common seasonal factors, we also show the gap between them (Kumamoto minus the six-prefecture average).

Source: Compiled by the HotelBank Editorial Team from MetroEngines Research (Kumamoto Prefecture N=436 properties / 19,623 rooms; other six prefectures N=2,327 properties / 128,051 rooms; July 2026 check-ins)

Now the numbers. For check-ins from 15 to 27 July (the 13 days before the quake), the gap averaged −1.07 points — Kumamoto consistently ran below the average of Kyushu’s other six prefectures. But for check-ins from 29 to 31 July (the three days after the quake), the gap averaged +4.76 points. That is a swing of 5.83 points across the event.

Source: Compiled by the HotelBank Editorial Team from MetroEngines Research (simple average of daily estimated OCC for July 2026 check-ins)

Extract the single sentence “immediately after a magnitude 7 quake, occupancy in the affected prefecture ran 4.8 points above neighbouring prefectures” and it could be read as a demand surge from an influx of recovery personnel. That reading is wrong — and the arithmetic proves it.

The properties classified as “B: listing continues, zero rooms remaining” in Section 2 numbered 19 properties and 1,851 rooms. Because their listings have not disappeared, they remain in the denominator of estimated OCC (total rooms), while the rooms whose remaining count went to zero are added directly to the numerator (sold rooms). The estimated-OCC population for Kumamoto Prefecture is 19,623 rooms, so 1,851 ÷ 19,623 = 9.4 points — the upper limit of the lift that category B alone can generate. The measured swing is 5.83 points, comfortably within the range that inventory withdrawal alone can explain. Add the partial contribution of the 1,292 rooms in “D: sharply reduced inventory days,” and the explanatory headroom is larger still.

There is one more discontinuity — in the observations themselves — that needs checking. The chart below plots the number of properties in Kumamoto Prefecture for which sales inventory for August check-ins could be observed, by observation date.

Source: Compiled by the HotelBank Editorial Team from MetroEngines Research (properties observed per observation date, August 2026 check-ins, Kumamoto Prefecture)

Between 20 and 28 July the observed property count moved between 304 and 394, then jumped to 493 on 29 July and 619 on 30 July, before returning to 340 on 31 July. The population itself is moving by nearly a factor of two from day to day. Running a daily comparison such as “occupancy on 30 July was higher than the day before” in this state measures a change in the observation set, not a change in demand. This is precisely why Section 2 fixed the population to “properties with at least 10 days of pre-quake inventory observation.”

7. Obon Demand Cannot Yet Be Judged

The obvious next question is what happens over Obon, and this article’s position is unambiguous: the current data cannot answer it. Here is the reasoning, in numbers.

The chart below compares estimated OCC for check-ins on 13 August 2026 (the Obon peak) between Kumamoto and Kagoshima prefectures, with lead time (LT) on the horizontal axis. LT15 at the right-hand end is 29 July — the day after the quake.

Source: Compiled by the HotelBank Editorial Team from MetroEngines Research (Kumamoto Prefecture 519 properties / 21,333 rooms; Kagoshima Prefecture 531 properties / 20,802 rooms; 13 August 2026 check-ins)

Three things can be read from it, and all of them point toward “not yet judgeable.”

First, there is only one post-quake observation point. The pre-aggregated curve this article references runs to 29 July (LT15), so effectively that single point is the only data after the quake (16:27, 28 July). Kumamoto’s estimated OCC rose 1.0 point, from 86.2% on 28 July 2026 (LT16) to 87.2% on 29 July 2026 (LT15), while over the same period Kagoshima fell 0.1 point, from 71.9% to 71.8%. That is a relative rise of 1.1 points — but you cannot state a trend from a single observation. And as Section 6 showed, movement in exactly this direction is what inventory withdrawal produces.

Second, the observed property count moved substantially on that one day. For Kumamoto’s 13 August arrival curve, the observed property count rose 7.8%, from 446 on 28 July to 481 on 29 July. You cannot discuss changes in the numerator while the denominator is moving.

Third, the curve contains thin cross-sections to begin with. The median observed property count on Kumamoto’s 13 August arrival curve is 396, but LT70 has only 21 properties, LT68 only nine and LT42 only 22. These are thin cross-sections falling well below 60% of the median, and this article excludes them from the chart (the breaks in the line are where they were removed). Plotted without exclusion, an aggregate from nine properties would connect to an aggregate from 396 properties with a line of identical weight, showing readers “troughs” and “peaks” that do not exist.

8. Demand Shifts to Other Kyushu Prefectures — What Became of the Asymmetry Argument?

The previous article presented a record from 2016: Oita Prefecture, some 87 km away, fell further (−39.4% year on year in May) than Kumamoto Prefecture, which contained the epicentre (−10.6% in May). The argument was that “seismic intensity decays with distance, but cancellations do not.” Can that argument be tested three days after the quake?

The short answer is that it cannot yet be tested, but the direction to monitor can be identified.

Look again at the chart in Section 6. For check-ins from 29 to 31 July 2026, Kumamoto’s estimated OCC (July 2026 check-ins) appears to rise, while the average of Kyushu’s other six prefectures falls from 86.0% to 84.5% to 80.1%. Kumamoto over the same period runs 90.4% → 90.0% → 84.5%. What can be observed at this point, in other words, is not a picture of “the affected prefecture rising as demand flows to neighbouring prefectures” but one in which “the affected prefecture rises in appearance through inventory withdrawal, while neighbouring prefectures actually fall.”

That said, this three-day decline cannot be attributed definitively to the earthquake. 29–31 July are Wednesday, Thursday and Friday, so comparison with the immediately preceding weekend carries a day-of-week factor. The fact that all seven Kyushu prefectures show a decline in the same direction also suggests common factors other than the quake are at work (within-week demand patterns, last-minute cancellations being reflected, and so on). In the 2016 record, the deepest declines across Kyushu’s prefectures came in the month after the event (May, for an event on 14 April) — one of the previous article’s lessons being that you must not judge from the figures of the month in which the disaster occurs. As for how far that 2016 disaster area subsequently recovered, our analysis Kumamoto Quake 10Y x Aso: Data on Recovery and Tourism Brand Rebound traces the recovery curve over a ten-year span.

It is worth setting out the indicators to monitor.

Source: Compiled by the HotelBank Editorial Team (2016 actuals per the previous article’s analysis based on the Japan Tourism Agency’s “Overnight Travel Statistics Survey”)

The rate baseline is also worth recording. According to MetroEngines Research, Kumamoto Prefecture’s estimated transacted ADR for August 2026 was approximately ¥12,200 (N=389 properties), against an actual of approximately ¥11,400 for August 2025 (N=385 properties) — a level of +7.0% year on year. Note, however, that this August figure was aggregated as of 26 July 2026 and does not reflect the 28 July earthquake. It is meaningful to record as a pre-quake baseline, but it does not represent current sales conditions.

9. Restoration and Reopening Announcements — “Coming Back” Information Began on Day 3

Listing only closure information misrepresents the situation. 30–31 July, the third day after the quake, was also the point at which reopening and restoration notices began to appear.

Source: Compiled by the HotelBank Editorial Team from MLIT, “On the Damage Situation Caused by the Reiwa 8 Kumamoto Earthquake (Report No. 12)” (as of 13:30, 31 July 2026), JR Kyushu and official announcements from each operating company

On the other hand, no lodging property within the scope of this survey has published a definite reopening date. Every property that announced a closure has used phrasing such as “for the time being” or “undecided” — understandable when water outages continue and there is no visibility on lifeline restoration. Whether Yatsushiro City’s roughly 25,300 households without water begin to be reconnected is the single biggest variable governing when the area can reopen.

10. Practical Implications for Operators, Municipalities and Publishers

To lodging operators: publish “we are open” as primary information on your own site. As this article has confirmed, the absence of inventory on online sales channels can be received by a third party as a signal that “this property may be closed.” Even when the reality is that the property is full with recovery personnel, or is simply throttling new sales, that distinction cannot be made from outside. Properties that posted their operating status with a date and time on their operator’s official site have been handled accurately even in third-party research like ours. The cost of posting is low and the effect in preventing misunderstanding is large. Notices that include an update timestamp (phrasing such as “*as of 09:00, 31 July 2026”) were particularly useful, because the freshness of the information can be judged.

To municipalities and tourism associations: do not build a “list of properties that are open” — build a route to primary information. As the previous article noted, a third party cannot guarantee that an individual property is operating. Our measurements substantiate that difficulty in numbers. The inventory data makes 36 properties look as though they have “disappeared,” but the number officially suspended is six within the verifiable range — and the top three properties whose inventory vanished were operating. What should be aggregated is not a determination of operating status but a route to each property’s official announcement. Beyond that, continuously publishing municipality-level lifeline information such as the number of households without water has practical value as material for inferring whether lodging is possible.

To those working with the data: do not use post-disaster occupancy as a demand indicator. As Section 6 showed, estimated occupancy in a disaster-hit area overshoots because of inventory withdrawal, and the indicator points in the opposite direction to reality. To treat a post-disaster area quantitatively, the minimum procedure is: (1) fix the population using pre-quake observation records; (2) check the observed property count daily and stop comparing if there is a discontinuity; (3) take the difference against an unaffected control area; and (4) look at actual volume in rooms rather than rates. This article followed that procedure — and still arrives at the conclusion that “Obon demand cannot be judged.”

To those involved in wide-area coordination: start from “rooms usable right now,” not total stock. The previous article mapped stock of 13,269 rooms within 30 km of the epicentre and 40,661 rooms within 80 km. This article indicates that roughly 8–15% of the stock within 30 km may already be unusable. Further, the example of the 651 rooms whose inventory went to zero suggests that a substantial number of rooms, though operating, are already filled by recovery personnel. Capacity planning must start not from peacetime stock but from what remains after subtracting two stages: stoppages caused by damage, and pre-emption by recovery demand. And that remainder increases rapidly by widening the distance band one step — a conclusion from the previous article that still holds.

11. Data Notes, Methodology and Limitations

Summary

In the 72 hours since the Reiwa 8 Kumamoto earthquake, transport has recovered substantially. The Kyushu Shinkansen resumed service between Hakata and Kumamoto on 31 July, expressway closures fell from 17 sections on 3 routes to 9 sections on 2 routes, and closures on nationally administered highways fell to zero. Neither Kumamoto Airport nor Amakusa Airfield has had a cancellation since 30 July. Water outages, however, still affect approximately 79,100 households, of which Yatsushiro City accounts for roughly 25,300. Just over 9,100 people remain evacuated, and 35 have died, including one case where the connection is still under investigation (both per Kumamoto Prefecture, 31 July).

On lodging supply, this article has presented two figures. Closure, suspension of operations or suspension of reservations could be verified through official announcements at six properties, 830 rooms; the inventory signal was lost in the data at 36 properties, 3,218 rooms. The reality lies between the two, and an order of magnitude of 1,000–2,000 rooms is the reasonable view. That corresponds to roughly 8–15% of the 13,269 rooms of stock within 30 km of the epicentre.

What this article most wants to say, however, is that the very fact that these two figures do not match is the lesson about how to read data during a disaster. The top three properties by room count whose inventory went to zero in the data — 651 rooms — were all operating normally according to their operators’ official announcements, and one of them had resumed online sales altogether by that same evening. Conversely, a property that officially announced it had suspended reservations through 30 September still showed 11 days of inventory in the data after the quake. The errors run in both directions.

That structure propagates straight into demand indicators. Kumamoto Prefecture’s estimated OCC (July 2026 check-ins) ran 1.07 points below the average of Kyushu’s other six prefectures over the 13 days before the quake, then 4.76 points above it over the three days after — a swing of 5.83 points. But this is not demand. The 1,851 rooms whose inventory went to zero represent 9.4% of the 19,623-room estimated-OCC population, creating up to 9.4 points of upward pressure on their own. Occupancy in a disaster-hit area reads higher the deeper the damage on the supply side. The indicator points in the opposite direction to reality.

The impact on Obon demand cannot be judged at this point. The booking curve for 13 August arrivals has only one post-quake observation point, the observed property count moved 7.8% on that single day, and the series contains cross-sections observed at just nine properties. In the 2016 record, the deepest declines across Kyushu’s prefectures came in the month after the event. Do not judge from the figures of the month of the disaster — that is the lesson carried forward from ten years ago.

We hope for the swiftest possible recovery in the affected areas.

Related Reading

• Kumamoto’s 11,702 Business Hotel Rooms: Can They Carry Recovery?
• Kumamoto Quake 2026: 2016 Recovery Curve as Kyushu’s 7-Pref Benchmark
• TSMC Fab 2 & Kumamoto’s Chip Corridor: Kikuyo–Koshi–Ozu ADR 2024–2026
• Kumamoto Quake 10Y x Aso: Data on Recovery and Tourism Brand Rebound
• Noto Earthquake +16 Months: Wakura Onsen Recovery & Summer 2026 ADR