São Paulo Heat & Transit Center

Official documentation states that the API supports authenticated line search, stops, vehicle positions, and arrival forecasts. Authentication requires a token generated in “Meus Aplicativos”. Official API docs

Provides current and hourly weather used in the heat map. These are modeled weather conditions, not street measurements. Documentation

Municipal pages state that the operation is activated when temperature or thermal sensation reaches 32°C or more. In January 2026 the city reported ten tents operating during qualifying heat days. Municipal reference

Published Mendeley Data V1 (17 Sep 2025), DOI 10.17632/6j7by525fy.1, CC BY 4.0. Covers 26,506 urban census tracts using Landsat 8/9 and other geodata. Open dataset

HEATLINK compares the pattern at each stop with the wider city and highlights unusual combinations of heat, local heat risk and service gaps. The result is an attention score, not a probability. The comparison is repeated so one random run does not decide the list.

Official municipal Favela boundaries. HEATLINK loads the published layer rather than drawing its own areas. Metadata

Official transit schedule used for stops and planned service. A timetable is never shown as an observed wait. Official source

SPTrans states that the municipal bus system has about 19.5 thousand stop points; approximately 6.5 thousand stops have shelter modules. HEATLINK uses this only to motivate the city-scale adaptation opportunity and never infers a specific stop's shelter status unless an authoritative stop-level inventory is supplied. Official stop-infrastructure page

Used only as the basemap.

HEATLINK samples city-scale apparent temperature across São Paulo so you can see how conditions change by place and time. It does not replace street-level heat measurements or personal exposure monitoring.

Journey heat load adds the time spent above your chosen heat threshold across walking, waiting, riding, transferring and the final walk. For formal city use, validate the measure against the local exposure framework you plan to use.

HEATLINK compares each stop using current heat, published local heat risk and the planned service gap. It repeats the comparison several times and combines the results. Passenger counts are added only when enough approved stop-level data is available. The result helps decide where to investigate first; it is not a health-risk probability, a causal estimate or a count of people affected.

When Favela focus is on, the list is limited to stops inside São Paulo’s official Favela map. HEATLINK does not invent population counts or turn unrelated administrative indicators into heat-vulnerability scores.

Suggestions are based on the information currently available for the selected stop or journey. They are planning prompts, not claims about health outcomes or guaranteed benefits.

The planner checks every combination of the actions you enter and finds the best mix within your budget, using the common benefit measure you choose. It does not guess costs, effects or interactions between actions.

If SPTrans is not connected, HEATLINK simply says live buses are unavailable. It does not add simulated vehicles.

The heat map cannot tell you the exact temperature under a particular shelter, tree or street canyon. Use local measurements for that.

HEATLINK does not estimate deaths, hospital visits or “people protected” without a locally validated health model and population data.

If the published heat-risk data or official timetable is missing, no priority list is shown.

The action planner starts blank. Outside studies are background information and are never treated as São Paulo effects automatically.

HEATLINK does not repurpose unrelated administrative scores as resident heat vulnerability.