Understanding Heat Risk
What is Heat Risk?
Heat risk describes the potential for heat to cause harm to people, animals, infrastructure, and essential services.
It reflects not only how hot it is, but also how factors such as humidity, duration of heat, night-time temperatures, and exposure increase the likelihood of health impacts and operational disruptions.
Heat risk is therefore different from temperature alone.
The same temperature can pose different levels of risk depending on local conditions and who or what is exposed.
How Heat Risk Is Calculated
Heat risk on this platform is derived from forecast weather data, including temperature, humidity, wind, and related heat-stress indicators.
To ensure clarity and consistency, heat risk is expressed using a common 0–4 scale across all sectors and departments.
While the same underlying weather data is used, different heat indicators are applied for different sectors to reflect how heat affects health, labour, animals, agriculture, electricity, education, and public services under Indian conditions.
Role of AI and Regional Downscaling
This platform leverages global AI-based weather forecast models as the starting point for predicting temperature, humidity, wind, and related variables. These global forecasts are then regionally adapted and downscaled to better represent local conditions across Karnataka.
Regional adaptation uses deep learning–based downscaling techniques, informed by local geographic and environmental factors such as elevation, land-use and land-cover, vegetation patterns, and ground-based observations. This approach improves representation of local heat conditions at district and taluk scales, where decisions are made.
Heat risk indicators are computed using scientifically established methods, while sector-specific risk levels and advisories follow guidance defined by the respective departments and local Heat Action Plans.
ℹ️ Regional information used for downscaling includes:
- Elevation and terrain
- Land-use and land-cover information
- Vegetation indices (e.g. NDVI)
- Ground-based weather station data (KSNDMC and IMD)
- Satellite-derived indicators (where available)
Heat Risk Levels (0–4)
| Level | Risk Category | Colour Code | Meaning |
|---|---|---|---|
| 0 | 🟢 No Risk | Green #2FA84F | Normal heat conditions. No heat-related impacts expected; regular activities can continue. |
| 1 | 🟡 Low Risk | Yellow #F2D21B | Mild heat stress possible, especially for sensitive groups. Stay hydrated and take basic precautions. |
| 2 | 🟠 Moderate Risk | Orange #E85C33 | Noticeable heat stress for vulnerable populations. Limit prolonged outdoor exposure and plan activities carefully. |
| 3 | 🔴 High Risk | Red #B0003A | High likelihood of heat-related illness. Avoid strenuous outdoor work during peak hours; cooling measures strongly advised. |
| 4 | 🟣 Severe / Extreme Risk | Deep Purple #4A003F | Dangerous heat conditions. Very high risk of serious health impacts; follow official advisories and emergency measures. |
- Day-time and night-time heat risks are calculated separately, as high night-time temperatures significantly increase health and labour risks.
- Department-specific advisories are aligned with these risk levels and with local Heat Action Plans (HAPs).
Role of Acclimatisation
Some heat indicators used on this platform assess heat not only against long-term local climate, but also against recent temperature conditions.
This helps identify periods when heat is unusually intense or arrives suddenly—situations known to increase health and labour risk due to limited acclimatisation.
Sector-wise Heat Risk Indicators
Different sectors experience heat stress in different ways.
The table below summarises the primary indicators used for each sector and the rationale.
| Sector | Primary Heat Indicator | Supporting Indicator | Why This Is Used |
|---|---|---|---|
| Health | Excess Heat Factor (EHF) + Night Heat | Heat Index (Maximum) | EHF captures heatwave severity relative to local climate, while high night-time temperatures increase mortality risk by preventing recovery. |
| Animal Husbandry | Temperature–Humidity Index (THI) | Tmax / Duration of high THI | THI is the established standard for assessing livestock heat stress in Indian conditions, especially for cattle and buffaloes. |
| Labour | WBGT (Maximum) | Heat Index / Tmax | WBGT is the global standard for worker safety, accounting for temperature, humidity, wind, and solar radiation—critical for outdoor and informal workers. |
| Agriculture | Tmax Anomaly / Heat Stress Days | Duration of hot spells | Crop stress depends on sustained temperature anomalies and consecutive hot days, not just single-day extremes. |
| Electricity | Tmax | Cooling Degree Days / Mean Heat Index | Electricity demand rises sharply with high temperatures and humidity due to increased cooling requirements. |
| Schools | Heat Index (Maximum) | Tmax | Heat Index reflects combined temperature–humidity stress and is commonly used in school safety and closure decisions. |
| Wind & Solar Energy | Tmax (Solar) | Low Wind Speed (Wind) | High temperatures reduce solar panel efficiency, while heat events often coincide with low wind affecting power generation. |
| Night Heat (Cross-cutting) | High Minimum Temperature | Consecutive Warm Nights | Elevated night-time temperatures amplify health and labour risks by preventing physiological recovery. |
How This Information Is Used
- Heat risk levels are updated using forecast weather data.
- The risk scale supports:
- Department-specific advisories
- Preparedness and response planning
- Coordination under Heat Action Plans
A common risk scale across sectors ensures clarity and avoids conflicting guidance.
Data and Data Sharing
ℹ️ Data Used
This platform uses ERA5 reanalysis data (ECMWF), IMDAA, IMD gridded datasets, and KSNDMC station and gridded data for heat-risk assessment.ℹ️ Forecast and Model Data Availability
Raw AI model forecast and hindcast data (2010–2025) for the Indian region are available on request.
Please contact climatehealth@artpark.in for access and collaboration.ℹ️ Research and Development Datasets
Heat risk indices such as Excess Heat Factor (EHF), Wet-Bulb Globe Temperature (WBGT), Heat Index, Tmax, and Wet Bulb Temperature (WBT) are available for 1991–present for Karnataka, to support further research, analysis, and method development. Please contact climatehealth@artpark.in for access and collaboration.
Limitations and Uncertainty
While this platform is designed to support heat-risk monitoring and early warning, users should be aware of the following limitations, which are common to all weather- and climate-based decision-support systems.
Forecast Accuracy
Heat risk assessments on this platform are based on AI-based weather forecasts that have been shown to perform comparable to, or better than, current state-of-the-art numerical weather prediction models (such as IFS and GFS), particularly in terms of spatial detail and short- to medium-range guidance.
However, forecast accuracy varies with:
- Lead time near-term (1-5 days) forecasts are more reliable than longer-range (6-15 days) forecasts, and
- Magnitude of extreme events (very intense heat events are inherently harder to predict precisely).
To improve reliability, bias correction and quantile correction are applied using available ground-based station data before regional downscaling.
ℹ️ Model Uncertainty
Like all forecast models, AI-based models may sometimes underestimate or overestimate heat conditions, especially during rapidly evolving or extreme weather situations.
This uncertainty is not unique to AI models and is inherent to all weather modelling approaches.Where forecast confidence is lower, the platform is designed to flag increased uncertainty, and users are advised to interpret risk levels with additional caution.
Sector-Specific Risk Calibration
Heat impacts vary across sectors, populations, and local contexts.
At present, comprehensive, calibrated thresholds linking heat indicators directly to sector-specific impacts (such as health outcomes, productivity loss, or crop damage) are limited or not uniformly available, particularly at local scales.
As a result:
- Sector-wise risk indicators are based on best available scientific evidence and operational practice, and
- Departmental advisories should always be interpreted in conjunction with local experience, vulnerability, and Heat Action Plans.
Use as Decision Support
This platform is intended to support preparedness, planning, and early action, not to replace official warnings or departmental decision-making processes.
Final decisions should continue to rely on official advisories, departmental protocols, and on-ground assessments.