Thailand: Housing and Flood Maps: What Can You Actually Verify?
Let’s explore this page together
Published · Updated
What a flood map can—and cannot—tell you about a home
A flood map becomes personal the moment a coloured patch appears beside a home you might rent or buy. Yet the patch may describe something quite different from the question you actually have: Was the garage flooded? Could residents reach the building? Did water reach the electrical switchgear? A map can establish a reported or modelled condition over a defined place and period. Unless its underlying evidence reaches inside the property, it cannot answer those building questions on its own.
There are several kinds of evidence in Thailand's official services, and their dates are not interchangeable. GISTDA, the country's space and geo-information agency, publishes satellite-based flood mapping and event reports. An October 6, 2026 report, for instance, describes imagery captured on October 5 over parts of Suphan Buri province. It tells readers where a particular observation was made; it is neither a continuous record of every wet street nor evidence that any named condominium's basement took on water.
The Department of Disaster Prevention and Mitigation (DDPM) offers a different perspective. Its risk dashboard draws on Tambon Smart Team surveys covering 2011–2020 and classifies areas using that historical assessment. The categories are useful for identifying places worth investigating, not percentages predicting whether an individual apartment will flood this year. ThaiWater, the National Hydroinformatics Data Center, deals in measurements. Its maps include 24-hour rainfall totals and water levels at monitoring stations.
Bangkok residents can also consult the Bangkok Metropolitan Administration (BMA). One official map is built around flood-risk and watch points documented in its 2022 action plan, including sites affected by heavy rain and drainage constraints. A separate road-flood monitoring system reports conditions at instrumented road locations and cameras. Both are local products. Their findings cannot be carried over to a house in Phuket, and a road sensor is not a sensor in the building's electrical room. Weather advisories form another category entirely. The Thai Meteorological Department's October 10, 2026 advisory forecast heavy to very heavy rain during October 10–13 in parts of Thailand; it did not certify flooding at every address named within a forecast region.
Finally, the water itself may come by different routes. River overflow can spread across a floodplain; an intense cloudburst may overwhelm a local drain; elevated coastal or downstream water can obstruct an outfall where conditions allow. GISTDA's work on urban flood-depth modelling adds a further, expressly modelled product in selected places, not a nationwide reading of the water height at every doorstep. The right question for the map is what it observed or calculated. What the building endured is the next question, for a different set of evidence.
Which map answers which question
Choose the evidence that matches your question. A satellite extent, a historical category and a road gauge may all be official, yet each measures something different.
| Data | Useful for | Cannot establish |
|---|---|---|
| GISTDA: observed flood extents | Mapped water during a named satellite observation | Indoor ingress or streets outside the observation |
| DDPM: area-level flood history | Area classifications from 2011–2020 surveys | A property probability or an actual building event |
| ThaiWater: rain and river gauges | Rainfall and water level at a monitoring station | Depth at an entrance or in a car park |
| BMA: 2022 flood-risk points | Bangkok locations flagged under the 2022 plan | Apartment ingress or today’s drainage condition |
| BMA: road-water stations | Readings and cameras at instrumented Bangkok roads | The condition of lifts, basements or electrical rooms |
| TMD: heavy-rain advisories | Areas and dates under a heavy-rain advisory | Actual inundation of a property |
Matching a property location to a map without false precision
A property's online pin is often only a convenient label for a development. That distinction becomes consequential when you move from browsing maps to making a housing decision. The front lobby, service entrance and underground car-park ramp need not face the same street or sit at the same elevation. In a detached house, the gate, the lowest part of the plot and the place where electrical equipment is installed may be more revealing than the centre of the roof. If the pin is approximate, the apparent distance to a flood boundary is approximate too.
The strongest starting point is an identifiable entrance or parcel supported by a site plan, building documentation or another reliable location record. A marketing address can take you to the neighbourhood; an official or verifiable plan tells you which side of the development you are testing. An automated map lookup can place a convincing-looking marker on the wrong plot. For example, floodwater beside a lower garage-access road matters even when it sits far from the development’s central pin.
Next comes the form of the flood information. A satellite-derived extent is a mapped interpretation of an image acquired at a particular time. Its edge has the spatial and detection limitations of that observation. A BMA point may label a flood-prone street location, while a DDPM shaded district carries an area classification. A ThaiWater marker represents a station measurement. Calling each of those marks an address-level flood boundary would collapse four different kinds of data into one misleading answer. A borderline overlap is best described as nearby or uncertain when the layer's accuracy has not been published.
Zooming in introduces its own trap. Modern basemaps can draw individual buildings sharply, while the flood overlay was created at coarser detail, from another date, or without published positional accuracy. A coordinate-reference label such as EPSG:4326 or EPSG:3857 specifies how positions are expressed. Neither label tells you the satellite's effective flood-detection detail. Nor can the size of a pixel be derived from the zoom level of the browser. If the source does not provide a defensible positional tolerance, a measurement that purports to show the flood edge a few metres from the door is precision without evidence.
Time deserves equal care. A date when water was observed, a publication date and today's date of access describe separate events. The BMA point map based on its 2022 plan does not become a 2026 street survey because someone opened the page in October 2026. Equally, an older recorded incident remains relevant evidence of what was reported at that time. A sound map record preserves those distinctions: which layer, which moment, which location marker, what its symbol means and which details of the premises have not been verified.
A map record you can use for one address
A map record you can use for one addressChecklist0 of 10
Why an unmarked area is not proof of a flood-free building
A blank space can look reassuring, especially when the map is official. The trouble is that official sources still have a defined purpose and a limited window on events. A satellite may pass after standing water has drained away, or before a cloudburst turns a narrow access road into a temporary pool. A dataset may have no observation for the date you care about. These are missing observations, not observations of dryness. The same caution works in the other direction: a station may record exceptional rain nearby without any evidence that the chosen address was inundated.
Urban streets present a particular technical challenge. In its June 30, 2026 explanation of a flood-prediction feature, GISTDA says synthetic aperture radar (SAR) can struggle to detect floodwater among tightly packed buildings because reflected radar signals interfere with the view of the water surface. That limitation is specific: radar mapping remains valuable with a suitable landscape, acquisition and processing method. A published flood polygon is a real piece of evidence about the area it depicts; an absent polygon is not a comprehensive negative test for a street hidden among buildings.
To address some of those gaps, GISTDA describes a separate urban modelling approach. It brings together terrain elevations, drainage and river geography, water-level and flow measurements, and satellite observations to simulate flood movement and potential depths. The agency describes pilot activity in Hat Yai and testing in Chiang Mai and Nakhon Si Thammarat, with broader expansion as an ambition. A digital elevation model helps represent slopes and low points; it is not itself a survey of the threshold of your chosen apartment building. A predicted depth needs a documented town, model run, scenario, time and vertical reference before it can inform any local discussion. It must not be presented as a universal depth layer already available and accurate at every Thai property.
A low category in a historical risk dashboard does not say that the ground floor has never been wet. The BMA's 2022 point descriptions offer a more concrete illustration: its flood-risk points refer to places susceptible under moderate-to-heavy rainfall, while monitoring points include places that warrant attention during more intense rain or where drainage is obstructed. Neither category certifies that water entered a particular flat.
Rain that accumulates in a local depression can take a different path from river water spreading over lowland. Near the coast, backwater or tidal conditions may reduce the ability of drains to discharge; they do not apply equally to every district. This is why a map's silence needs to be followed by building evidence, while a map's positive finding needs to be translated into a precise question about access and equipment—not an automatic verdict on the home.
Misreadings that can change a housing decision
No mapped flood polygon means the building stayed dry.
SAR can miss urban water among buildings; acquisition time also matters.
TipTreat blank coverage as inconclusive and look for dated building records.
An area-level risk label means floodwater entered this apartment.
DDPM classifies areas using 2011–2020 evidence, not the interior of a unit.
TipInvestigate the actual building before accepting or rejecting it on that label.
The river gauge tells me how deep the water was at the entrance.
ThaiWater gauges report conditions at their stations; elevations may use different references.
TipDo not turn a river-stage number into doorstep flood depth.
A forecast depth layer works at parcel scale everywhere in Thailand.
GISTDA describes modelled depths for selected urban tests, subject to model assumptions.
TipEstablish the model extent and status before using it for this property.
Evidence from the building can change the picture
The practical stakes of flooding rarely stop at the front door of an apartment. An upper-floor resident might have a dry living room while a flooded approach blocks the car park, a lift is taken out of service, or essential equipment has to be shut down. Conversely, a street photographed under water does not prove that the development's pumps failed.
For a condominium, the building manager and condominium juristic person are natural custodians of operational records. Dated maintenance logs, reports of pumping, resident notices, repair invoices or redacted incident records can link an event to a particular entrance, garage level, lift pit or plant room. A generic reassurance that the neighbourhood does not flood is much weaker, especially if the person giving it has never had access to the building's records.
Drainage design should account for runoff entering a low garage ramp or a backed-up outlet. A drawing showing the intended system is not proof of its performance under an actual storm. Likewise, evidence of an old failure should be read alongside later alterations: raising an entry sill, relocating equipment or fitting barriers might change the consequences, but only if the measures actually exist and their operation is understood. No pump or barrier can be treated as an unconditional promise of a dry building.
Elevation comparisons are particularly vulnerable to false confidence. Street surface, garage floor, threshold, electrical-room floor and historic high-water mark are separate measurements at separate places. Thai government open data describe Flood Mark observations collected from visible traces on posts, walls or fences, with heights expressed in metres relative to mean sea level. That is a recorded water-surface elevation, not an apartment-level flood depth. To calculate how high water stood above a specific threshold, the threshold elevation and water mark would need compatible units, a common vertical datum, a relevant location and matching event dates. Subtracting a nearby river-stage figure from an elevation lifted from an unrelated terrain map would produce a number without a sound physical meaning.
Not every manager will release internal documents, and legitimate privacy or record-keeping limits may apply. Redacted reports and site plans can still clarify a problem without disclosing residents' identities. If the only answer available is a verbal statement, uncertainty remains; unavailable records are not themselves evidence of concealment. Insurance deserves a parallel review of actual wording, including exclusions and whether damage to shared equipment is treated differently from loss within a unit. For a long-term tenant, interrupted access and utilities may be the immediate concern. A purchaser also needs to understand who pays for common-area repairs and whether past damage has been remedied. Those are property questions a national map cannot settle.
Questions for building management before renting or buying
Questions for building management before renting or buying
- Have roads, entrances, parking levels or plant rooms flooded? Can you share dated reports or photographs?
- Is there a drainage plan, pump service log and evidence of performance during heavy rain?
- What are the road, ramp, entry and electrical-room levels, and do they share a documented vertical datum?
- During previous events, were access, electricity, water supply or lifts interrupted?
- Which flood barriers, backflow valves and pumps have been installed and tested, and what is the emergency procedure?
- Which kinds of flood damage are covered for the building and unit, and what exclusions apply?
Making a decision when the map and building evidence disagree
When the sources have been read, the most useful outcome is an evidence statement about the home—not a flood-risk score made up from colours. One is that a documented event affected the surrounding area, but entry of water into the chosen building is unconfirmed. To establish the second, you would still need the condition of the access road, threshold, garage and technical rooms at the relevant time.
Another outcome is direct evidence from the premises. A dated garage incident report identifying the building, backed by repair or maintenance records, changes what can reasonably be said. It may establish a past ingress or service interruption, even though the current condition of the repaired systems still requires attention. The reverse is equally important: an authenticated record that a monitored entrance remained dry during a particular event is valuable evidence for that place and time, but cannot guarantee the result of a different storm. A single dramatic photograph may have less explanatory value than an unremarkable document that gives the location, date and actual impact.
The third outcome is a coverage gap. A Bangkok-only point map cannot test a Phuket address; a 2011–2020 historical dashboard cannot tell you what happened after a newer development was built. An expired weather forecast or a satellite image acquired outside the event window has the same basic limitation. If the precise entrance is unknown, even a highly detailed regional layer cannot be linked to that entrance credibly. 'No information for this property' should never be edited into 'no flood history.'
Conflicting evidence is the fourth outcome. Apparent disagreement often disappears once dates and definitions are separated. A road sensor measures the street where it is installed; a manager may be talking about a dry lobby. If a dated photograph of an entrance does genuinely contradict a report from the same building and event, however, the sensible next move is to establish provenance and seek original records. Both may be accurate about different parts of the site.
What you do with the result depends on exposure and the consequences you could accept. Temporary loss of one road with an alternative route is a different housing problem from verified damage to communal power equipment with unclear repairs. Insurance terms, access arrangements, present drainage protection and responsibility for shared works can all change the decision. For a purchase with material unanswered questions, an independent local engineering assessment may be justified; it has not been performed by this article or by the public services cited here. For a rental, a dated account of access and utility interruptions can be enough to expose a serious day-to-day drawback. In either case, keep the conclusion attached to the evidence and its date. No historical map offers a guarantee about what the next flood season will do.
Expert view

A high-floor apartment can still become difficult to live in when access and shared services fail. I pay particular attention to the garage ramp and the location of critical electrical equipment. Water often has a route into a site that an area-level map cannot reveal. That route matters more to the housing decision than the colour attached to the neighbourhood. Records of past interruptions and subsequent repairs can give a building a more meaningful history. Pumps and barriers are useful only when they are properly installed, maintained and tested. A blank satellite map should never be the only reassurance offered to a buyer or tenant.
Sources and check dates
Sources5
- GISTDA — Flood | Disaster Platform
Official flood interface with province, district and subdistrict filters; an individual layer still needs its acquisition date and scale checked.
- GISTDA — Flood Situation
Event-based satellite news with specific dates and locations, not continuous property monitoring.
- GISTDA — THEOS-2 flood imagery over parts of Suphan Buri, 5 October 2026
THEOS-2 image acquired October 5 and published October 6, 2026 showing affected parts of Suphan Buri; not evidence about any named condominium.
- GISTDA — Flood Area Prediction System and urban SAR limitations, 30 June 2026
Agency explanation of urban SAR blind spots, terrain-based flow modelling and selected-city tests; not proof of nationwide parcel-level depths.
- DDPM — Flood Risk Dashboard
Historical classifications grounded in Tambon Smart Team surveys from 2011–2020, not apartment-specific flood probabilities.
Useful next reading
A practical second opinion
Need help assessing a specific property?
Send the property link, price and your goal. We will separate the documented facts from the points that still need a property-specific check.
What to send: link, budget and target date