Heat Flow — How Warmth Moves Through Tokyo
Shinjuku doesn't just heat itself. It heats Koto, 8 kilometers away, by 1.5 degrees Celsius on a still August night. It heats Setagaya through a slow westward crawl that takes six hours to complete. It pushes warmth south toward Shinagawa until the bay breeze pushes back. And it fails to heat Nerima at all — blocked by a ridge in Mejirodai that creates one of the sharpest thermal fronts in the 23 wards.
If you want to understand Tokyo's weather, you don't start with the forecast. You start with the heat. And the heat starts in Shinjuku.
The Skyscraper Effect
Shinjuku's heat engine has three cylinders. The first is the skyscraper wall. The Tokyo Metropolitan Government Building, the Mode Gakuen Cocoon Tower, the dozen other structures over 150 meters — these buildings compress warm air at street level and create vertical circulation cells that pull heat upward during the day and trap it at night. The canyon effect is well-documented in urban climatology, but Shinjuku's specific geometry — a dense cluster of towers surrounded by lower residential fabric — creates a particularly strong version of it.
The second cylinder is air conditioning exhaust. Tokyo's commercial buildings dump waste heat into the streetscape at a rate that peaks between 2pm and 5pm, precisely when solar loading is also at its maximum. A single large office tower can exhaust enough heat to raise the ambient temperature of the surrounding 200 meters by 0.3 to 0.5 degrees. Multiply that by Shinjuku's tower count and you're looking at a significant artificial heat source layered on top of the solar loading.
The third cylinder is traffic density. Shinjuku Station handles over 3 million passengers daily. The bus terminal, the taxi queues, the delivery vehicles servicing the commercial core — all of them add engine heat, brake heat, and exhaust heat to the local energy budget. On a weekday afternoon, vehicle heat alone contributes an estimated 0.2 to 0.4 degrees to the Shinjuku microclimate.
Add these three together and you get a district that runs 2 to 3 degrees hotter than the JMA baseline for "central Tokyo" on a clear summer day. And that 2 to 3 degrees doesn't stay in Shinjuku. It flows.
Eastward: The Bay Release
The strongest heat flow from Shinjuku heads east. This is the path of least resistance — flat terrain, few obstructions, and a thermal sink at the end in the form of Tokyo Bay. Hot air rises over Shinjuku, creates a localized low-pressure zone at street level, and draws in slightly cooler air from the surrounding districts. But the overall net movement is outward, and the eastern corridor toward Chiyoda and then Koto is the widest channel.
By 3pm on a typical August day, the thermal plume extending east from Shinjuku is detectable in temperature readings from Suidobashi and Ochanomizu — districts that should, by latitude and elevation, be virtually identical to Shinjuku. They're not. They're 1 to 1.5 degrees warmer than they would be without Shinjuku's influence. By 6pm, that plume reaches Koto, where the bay water begins to absorb it.
Koto's role in this system is as a thermal release valve. The shallow bay water warms quickly under direct solar loading, but it still has enormous heat capacity compared to asphalt. When the eastward heat flow reaches Koto's waterfront, the energy transfers into the water column rather than continuing to warm the air. This is why Koto often feels milder than Chiyoda at 7pm — not because Koto is cool, but because it's where the heat finally finds somewhere to go.
On thermal satellite imagery, you can see this eastward plume as a warm tongue extending from central Tokyo toward the bay. It's most visible at 9pm, when the heat island has peaked and the bay is still absorbing. The plume's width varies with wind speed — on still nights it's narrow and intense, on breezy nights it diffuses across a broader front. But it's almost always there.
Westward: The Residential Absorption
Heat also flows west from Shinjuku, but this flow behaves very differently. The terrain rises gradually toward the Tama Hills, and the land use shifts from commercial towers to residential streets with more vegetation. Setagaya is the primary recipient of this westward heat flow, and it absorbs the energy differently than Koto does.
Where Koto releases heat to the bay, Setagaya stores it. The residential fabric of Setagaya — low-rise buildings, tree-lined streets, pocket gardens — has higher thermal mass and lower heat conductivity than Shinjuku's concrete canyons. Heat that arrives in Setagaya at 4pm doesn't bounce back. It sinks into walls, soil, and pavement, then releases slowly through the evening.
This creates a delayed warming effect. Setagaya often reaches its peak temperature two to three hours after Shinjuku does. If you cycle west from Shinjuku at 2pm, you'll feel cooler air as you leave the tower district behind. But if you make the same trip at 7pm, you'll find that Setagaya has warmed significantly while Shinjuku has begun to cool. The 3 degree drop that cyclists feel heading west from Shinjuku at 7pm is real — but it's the result of Shinjuku cooling faster, not Setagaya staying cool.
Komazawa Olympic Park plays a surprising role in this westward flow. The park's large green space creates a cool pool that interrupts the heat flow. Air temperature measurements at the park's center can run 2 degrees lower than the surrounding residential streets on a summer afternoon. This cool pool acts as a partial barrier, deflecting some of the westward heat flow north toward Shimokitazawa and south toward Futako-Tamagawa. It's one reason why the thermal gradient between Shinjuku and Setagaya isn't linear — there's a cool dip in the middle.
Southward: The Sea Breeze Pushback
The southern heat flow from Shinjuku is the most contested. Hot air heads south toward Shinagawa, but the bay breeze — when it's active — pushes northward up the same corridor. The result is a thermal front that shifts position throughout the day.
On strong sea breeze days, the front can reach as far north as Shibuya by early afternoon. This means the southern heat flow from Shinjuku gets blocked, deflected eastward toward Minato or westward toward Nakano. You can sometimes stand on the platform at Shibuya Station and feel the temperature difference between the north and south exits — the south exit is in bay breeze territory, the north exit is still in Shinjuku's heat plume.
On weak sea breeze days, Shinjuku's heat pushes all the way to Shinagawa. The district becomes an extension of the heat island, and the bay breeze doesn't arrive until late afternoon or not at all. These are the days when Shinagawa feels unexpectedly hot — not because Shinagawa has changed, but because Shinjuku's heat has reached it before the cooling sea air could.
The boundary between these two regimes is surprisingly sharp. A difference of 2 kilometers per hour in the sea breeze speed can shift the thermal front by a full kilometer inland. This is why cyclists on Route 246 — the main road connecting Shinjuku to Shinagawa — often report sudden temperature changes at unpredictable locations. They're crossing and re-crossing a front that moves faster than they do.
Northward: The Blocked Front
The northern heat flow from Shinjuku is the most dramatic — because it mostly fails. The Mejirodai ridge, a low east-west elevation running through Toshima and Nerima wards, blocks the northward movement of warm air. The ridge is only 30 to 50 meters high, but that's enough to create a thermal barrier on still days.
Hot air rising over Shinjuku creates a weak circulation cell that draws in cooler air from the north at ground level. This means the net heat flow northward is actually negative — Shinjuku is pulling cool air down from Nerima rather than pushing warm air up toward it. The result is a thermal front along the ridge line where temperatures can drop by 3 to 4 degrees over a distance of less than 2 kilometers.
This front is most visible in the early evening. If you take the Seibu-Shinjuku Line north from Takadanobaba to Nerima between 6pm and 8pm on a summer day, you can feel the temperature drop as the train crests the ridge. It's not air conditioning in the tunnel. It's a real thermal boundary that the railway happens to cross.
Nerima residents know this boundary well. They'll tell you that their ward has "proper seasons" compared to central Tokyo, and they're not wrong. Nerima's connection to the outside weather system — through the Tama River gap and over the Mejirodai ridge — means it experiences a climate that's genuinely different from Shinjuku's. The ridge is the wall that separates them.
Nighttime Heat Release
The heat flow doesn't stop when the sun goes down. In some ways, it becomes more important. Shinjuku's concrete and asphalt — heated through the day — continues to emit thermal energy through the night. The peak heat island effect in most Tokyo districts occurs not at noon, but between 9pm and midnight.
This nighttime heat release creates a secondary eastward flow that's actually stronger than the daytime version in some respects. Without solar heating to drive convection, the circulation is slower but more persistent. The heat plume from Shinjuku drifts eastward at walking pace, reaching Koto's waterfront by 11pm or midnight. You can see it in the temperature data — Koto's nighttime minimum is often 1 to 2 degrees higher than it would be without Shinjuku's delayed heat contribution.
The concrete continues emitting until 2am on a typical summer night, and some surfaces — particularly the dark asphalt of large parking structures — are still releasing stored heat at 4am. This is why Tokyo's heat island is so persistent. It's not just the daytime generation. It's the 14-hour release cycle that extends thermal influence deep into the night.
For cyclists and night workers, this nighttime flow matters. The eastward corridors from Shinjuku — through Iidabashi, Suidobashi, and into Koto — stay noticeably warmer than the surrounding districts until after midnight. Meanwhile, the westward flow into Setagaya has dissipated by 10pm because Setagaya's lower thermal mass releases its stored heat faster. The city has two different thermal personalities at night, and they're both determined by where Shinjuku's heat went during the day.
The Cyclist's Perspective
All of this is invisible until you move through it. A person who stays in one district experiences their local climate as "the weather." But a cyclist crossing the city at 7pm experiences Tokyo as a thermal landscape — rising and falling, pooling and flowing, blocked by ridges and released by water.
The 3 degree drop heading west from Shinjuku at 7pm is the most commonly reported gradient, but it's not the only one. Heading east from Shinjuku at 9pm, you'll feel the temperature rise as you enter the heat plume's path, then fall again as you approach the bay. Heading south from Shibuya at 11am, you'll hit the sea breeze front somewhere between Gotanda and Osaki — the exact location varies, but the sensation is unmistakable. And heading north from Ikebukuro toward Nerima at any time of day, you'll cross the Mejirodai thermal front if the wind is calm.
These gradients shape behavior in ways that most residents don't consciously register. People choose their evening walking routes around known cool corridors. Convenience store placement follows the heat island map — stores in heat pockets sell more cold drinks after 8pm. Even cherry blossom viewing patterns are influenced, with the warm plume districts blooming 2 to 3 days earlier than the cool edge districts in a typical spring.
Tenki Link exists to make these flows visible. The temperature differences you see on our network map are the connections between districts that most weather forecasts ignore. Shinjuku heats Koto. Setagaya absorbs and delays. Shinagawa pushes back. And Nerima stays separate, behind its ridge, feeling the outside weather while the city bakes. It's all one system. You just have to ride through it to feel it.