Skip to content
LIVEHeat-stress early warning · India

See the heatwave coming. Ward by ward.

Thermora turns NASA POWER climate data and local vulnerability into a five-day heat-stress forecast for every ward, so city teams and hospitals can act before temperatures peak.

Explore the heat mapRequest a pilot
Live · Ghaziabad, UPDemo data
HTSI
78
Danger
Heatwave
82%
probability
Hospital load
91%
bed use

Ward-level heat map

Heat never hits a city evenly. Neither should the response.

Thermora scores every ward on the same 0 to 100 heat-stress scale, so teams can see which neighbourhoods cross into danger first and send water, shade and medical support there.

Ghaziabad ward risk

Heat-stress index per ward. Hover or tap a ward for details.

LIVE
HINDONCityKavi NagarVijay NagarMohan NagarVasundhara
CoolExtreme
City 78
020406080100
Demo dataIllustrative ward layout

Hottest wards

Wed · Day 3
  1. 01

    Ward 19 · City

    88
  2. 02

    Ward 22 · Mohan Nagar

    88
  3. 03

    Ward 18 · City

    86
  4. 04

    Ward 23 · Mohan Nagar

    86
  5. 05

    Ward 20 · City

    85
  6. 06

    Ward 34 · Mohan Nagar

    85

Zone averages

  • CityExtreme81
  • Kavi NagarDanger75
  • Vijay NagarDanger77
  • Mohan NagarExtreme80
  • VasundharaDanger78

City average 78 on Wed. Demo data.

Heat-stress index

One number for how dangerous the heat really is.

Temperature alone misses the full picture. HTSI blends heat, humidity, sunshine, still air and who lives in each ward into a single 0 to 100 score with four clear bands.

HTSI at the forecast peak

Ghaziabad city average

LIVE
020406080100SAFECAUTIONDANGEREXTREME

78/100

DANGER

Wed, Day 3. Peak 46.8 °C, night low 32.4 °C, heatwave probability 82%.

  • Safe0–39
  • Caution40–59
  • Danger60–79
  • Extreme80–100
Demo data

What is driving the score

Share of HTSI 78 by factor

LIVE

  • Temperature

    T2M_MAX · 46.8 °C

    34%

    Afternoon maximum at 2 m

  • Humidity

    RH2M · 38 %

    21%

    Raises how hot the air feels

  • Solar radiation

    ALLSKY_SFC_SW_DWN · 7.6 kWh/m²

    18%

    Sun load on streets and roofs

  • Wind

    WS2M · 1.8 m/s

    9%

    Still air keeps the heat trapped

  • Vulnerability

    Ward layer · 0.71 index

    18%

    Who lives here and how exposed they are

Bars are scaled to the largest driver. Wind counts through its absence: on still days the heat stays trapped in the streets.

Demo dataInputs: NASA POWER parameters and ward data

5-day forecast

Five days of warning, not five hours.

An ensemble forecast tracks air temperature every three hours, with a confidence band that widens as the horizon grows, and flags the day the heatwave peaks.

5-day heatwave forecast

Air temperature every 3 hours with its confidence band

LIVE
  • Air temperature
  • Confidence band
  • 45 °C heatwave line
  • Forecast peak

  • Mon

    43.6°

    54%

  • Tue

    45.2°

    71%

  • Wed

    46.8°

    82%

  • Thu

    45.9°

    76%

  • Fri

    43.1°

    48%

Five-day forecast, demo data
DayMaximum °CMinimum °CHeatwave probabilityHTSI
Mon43.629.854%66
Tue45.231.271%73
Wed46.832.482%78
Thu45.932.076%75
Fri43.130.648%64
Demo dataGhaziabad · ensemble output

Temperature anomaly

Daily maximum against the seasonal normal, March to June

LIVE
MarAprMayJunMonWedFri
Cooler than normalExtreme heat
−4 °C−1+2+5+8 °C
Days at +4.5 °C or more
9
Longest hot spell
5days
Peak anomaly
+5.9°
15 May, 46.5 °C
Demo dataDaily heat intensity

Hospital readiness

See the surge before it reaches the emergency ward.

Heat-related admissions rise with the heat. Thermora turns the forecast into expected admissions and bed utilisation, so hospitals can add staff and free beds before the capacity line is crossed.

Hospital surge outlook

Expected heat-related admissions and district bed utilisation

LIVE
  • Admissions, observed
  • Admissions, forecast
  • Above capacity
  • Bed utilisation (right axis)
  • 85% capacity threshold
Hospital surge outlook, demo data
DayPhaseHeat-related admissionsBed utilisation
−4dobserved3872%
−3dobserved4174%
−2dobserved4676%
−1dobserved4475%
Todayobserved5279%
Monforecast6383%
Tueforecast7887%
Wedforecast9691%
Thuforecast8889%
Friforecast6184%
Demo dataBars read on the left axis, the line on the right

Peak bed utilisation

91%

Forecast for Wed, Day 3. Demo data.

Admissions on the peak day

96

Up 85% on today. Demo data.

Days above 85%

3days

Tue to Thu. Demo data.

Warning lead time

2days

Before the capacity line is first crossed. Demo data.

Ensemble model

Three models. One forecast.

Each model sees the heat differently. Thermora blends them into one weighted ensemble, so no single model's blind spot becomes the city's.

Ensemble contribution

Share of the final forecast by model

LIVE

3

models, one forecast

  • XGBoost45%
  • Random Forest30%
  • LSTM25%
Demo dataWeights shown are illustrative

XGBoost

Gradient-boosted trees

Reads daily weather and ward features and catches the sharp, non-linear jumps in heat stress.

Random Forest

Bagged decision trees

Hundreds of trees vote together, which keeps the forecast steady when the inputs are noisy.

LSTM

Sequence neural network

Follows how heat builds up day after day, so multi-day heatwave spells are not missed.

Each model produces its own five-day outlook. The ensemble weights them into a single HTSI and heatwave probability for every ward. The weights shown here are demo values.

How it works

From satellite to street in four steps.

Thermora connects global climate data to the streets where people actually feel the heat, then turns it into decisions teams can act on.

  1. 01

    Ingest

    Daily climate data arrives from NASA POWER: maximum temperature, humidity, solar radiation and wind over the city.

    • T2M_MAX
    • RH2M
    • ALLSKY_SFC_SW_DWN
    • WS2M
  2. 02

    Localise

    Ward layers add who is exposed: elderly residents, housing density, green cover and access to cooling.

    • Population
    • Housing
    • Green cover
  3. 03

    Forecast

    An ensemble of XGBoost, Random Forest and LSTM projects HTSI and heatwave probability five days ahead.

    • 5-day
    • 3-hourly
    • Ensemble
  4. 04

    Act

    Risk bands, hospital surge outlooks and peak-day flags give every team a clear next step, ward by ward.

    • Wards
    • Hospitals
    • Peak day

Request a pilot

Bring ward-level heat forecasts to your city.

Tell us about your city and your team. We will walk you through Thermora with your own wards and your own heat season.

  • Municipal corporations

    Prioritise cooling centres, water points and ward teams before the hottest days arrive.

  • Hospitals and health departments

    Plan staff, beds and supplies for the surge days ahead.

  • Disaster management teams

    Trigger heat action plans with ward-level evidence behind every call.

We only use these details to reply to your request.

Made with Modulify