About Ama Sora — Sea Breeze Observers

Ama Sora started with a bicycle and a question. Marina Kondo was riding the Tokyo Bay coastline in August 2019, training for a long-distance cycling event, and she noticed something that didn't match the weather forecast. The app on her phone said 33°C and calm wind. Her body said 29°C and a steady breeze from the bay. The forecast was technically correct — it was 33°C at Shinjuku, where the forecast model was centered — but it was wrong where she was. The sea breeze had created a pocket of cooler air that no forecast she could find was capturing accurately. She started looking into it, and the more she looked, the more she realized how little public information existed about one of Tokyo's most reliable natural cooling systems.

Marina Kondo — Founder & Lead Researcher

Marina holds an MSc in Physical Oceanography from the University of Tokyo, where her thesis focused on the sea-breeze circulation over the Seto Inland Sea. She grew up in Kagawa Prefecture, on the north shore of the Inland Sea, where the summer sea breeze is so reliable that local farmers set their clocks by it. When she moved to Tokyo for graduate school, she expected the bay breeze to be a weaker version of the same phenomenon. It is — but it's also more complex, more compressed, and more contested by the urban environment than anything she studied in the relatively rural Inland Sea region.

In the summer of 2021, Marina cycled the entire Tokyo Bay coastline — all 180 kilometers from Choshi to Uraga — with a handheld anemometer mounted on her handlebars, logging wind speed and direction every minute. The dataset she collected, combined with temperature loggers at six inland points, forms the backbone of Ama Sora's observational record. She has since published two peer-reviewed papers on Tokyo's sea-breeze front dynamics and presented her findings at the Japan Meteorological Society's annual meeting in 2023.

Marina's approach combines rigorous physical oceanography with a stubborn insistence on field validation. She doesn't trust model output until she's stood at the location and felt the wind on her face. She's been wrong enough times — models predicting breeze when none came, or missing a strong front that wasn't in the forecast — to maintain a healthy skepticism of any data she hasn't personally verified. This philosophy permeates Ama Sora. Every claim we make is backed by either direct observation or transparent, reproducible analysis of public data.

When she's not cycling the bay or staring at satellite imagery, Marina works as a freelance oceanographic consultant for offshore wind energy projects. She splits her time between Jersey City, New Jersey (where Ama Sora Research LLC is registered) and Toyosu, Tokyo (where the field station operates). You can reach her at research@ama-sora.com.

Tetsu Yamamoto — Co-Founder & Meteorologist

Tetsu spent 22 years as a weather officer at the Japan Meteorological Agency's Haneda Airport office, where he tracked sea-breeze conditions for aviation forecasting. Haneda is one of the best observation points in Tokyo for the sea breeze — right on the bay shore, with unobstructed exposure and a long, continuous data record. Tetsu estimates he's personally verified over 800 sea-breeze days during his tenure, comparing model output against actual surface conditions at the airport.

His expertise is in the practical meteorology of the sea breeze — the subtle cues that distinguish a strong-breeze day from a weak one, the synoptic patterns that enhance or suppress the circulation, and the forecasting challenges that make sea-breeze prediction one of the hardest problems in local meteorology. He has a particular interest in the interaction between the sea breeze and the urban heat island, which he calls "the frontal battle" — the daily summer contest between cool marine air and hot urban convection.

Tetsu retired from JMA in 2022 and joined Ama Sora full-time in 2023. He brings institutional knowledge that can't be found in published papers: the historical evolution of JMA's sea-breeze forecasting methods, the undocumented quirks of the Haneda observation site, and the oral history of notable sea-breeze events from the perspective of the forecasters who tried to predict them. He's also our primary liaison with JMA, facilitating access to AMeDAS historical data and MSM model output for our research.

Sora Ichikawa — Contributor & GIS Specialist

Sora is a GIS and data visualization specialist who joined Ama Sora in 2022. Her contribution is the visual infrastructure of the project — the bay-breeze front map on the home page, the temperature-distance profile on the coastal wards page, and the data processing pipelines that turn raw observations into the displays you see. She built the SVG front map from Tokyo's 3D city data, using building height models to estimate aerodynamic roughness and wind channeling effects across the coastal wards.

Sora's background is in urban planning, not meteorology, and she brings a valuable cross-disciplinary perspective. She's interested in how the sea breeze interacts with the built environment — how a new high-rise can block the breeze for a neighborhood, or how a green corridor can channel it inland. She's currently developing a micro-scale wind model for the Ginza-Shimbashi corridor that combines our field measurements with computational fluid dynamics simulations of building-scale airflow. The goal is to produce street-level sea-breeze forecasts that account for the specific geometry of Tokyo's urban canyons.

Sora works remotely from Sapporo, which she says gives her a useful outsider's perspective on Tokyo's microclimate. "In Sapporo, we don't have a sea breeze — we have a mountain breeze," she notes. "The comparison is interesting. Both are thermally driven local circulations, but the mountain breeze is more reliable because the terrain doesn't change, while Tokyo's sea breeze is constantly fighting the city."

What We Do

Ama Sora operates on three levels. First, we maintain a continuous observational record of Tokyo Bay sea-breeze conditions during the warm season (April through October). This involves daily monitoring of AMeDAS data, weekly field visits to our instrumented transect, and monthly analysis of satellite imagery. Second, we publish our findings — on this website, in peer-reviewed journals, and at scientific conferences — with a commitment to open data and reproducible analysis. Third, we provide practical guidance to cyclists, urban planners, and anyone who wants to use the sea breeze to make their summer in Tokyo more comfortable.

We don't sell weather forecasts. We don't claim to predict the future. What we do is observe, document, and explain — with the rigor of scientists and the grounded practicality of people who spend their afternoons on bicycle seats, feeling the wind change direction at the frontal boundary. If our work helps you choose a cooler cycling route, understand why your apartment is hotter than your friend's across town, or simply appreciate one of Tokyo's underappreciated natural phenomena, then we've done our job.

Our Equipment

Our field kit is intentionally simple. We use Onset HOBO U23 Pro v2 temperature loggers (±0.2°C accuracy, 1-minute logging interval) for the instrumented transect. Wind measurements come from a Kestrel 5500 handheld weather meter mounted on Marina's bicycle handlebars. For upper-air data, we rely on JMA radiosonde launches from Tateno (Ibaraki) and the Himawari-8 satellite's 10-minute full-disk imagery. The total cost of our portable equipment is under $2,000 — we believe good science doesn't require expensive instruments, just careful methodology and consistent fieldwork.

Our data processing uses open-source tools: Python with xarray and pandas for data analysis, GDAL and QGIS for spatial processing, and D3.js for web visualizations. All code is available on request, though we haven't yet set up a public repository (it's on our to-do list). The website itself is hand-coded HTML, CSS, and JavaScript — no frameworks, no tracking, no analytics except the cookie-based topic counter that you can decline.

Why "Ama Sora"

The name comes from the Japanese ama (海, sea) and sora (空, sky). It's a simple description of what we study — the interaction between the sea and the sky above Tokyo Bay, and the wind that connects them. In some Japanese dialects, particularly in coastal Kagawa where Marina grew up, ama also refers to the women who traditionally dove for shellfish and seaweed — people who worked at the boundary between sea and land, intimately familiar with the daily rhythm of wind and tide. We like that resonance. We're not divers, but we do work at a boundary, and we do pay attention to the daily rhythm of a force that shapes the city whether its residents notice it or not.