Heatwave events are primarily driven by temperature fluctuations. To identify potential severe heat events, we selected land surface temperature (LST) as the primary metric. The baseline dataset comprises VIIRS global LST data to calculate anomalies relative to a 10-year climatology. MODIS global LST serves as a backup for analyzing anomalies over a 20-year climatology when needed. Severe events, including their boundaries and severity, are detected based on the calculated mean temperature anomaly. Detection is carried out on both daily and monthly scales. While daily analysis captures short-term event dynamics, monthly analysis provides a broader perspective on annual trends and potential correlations with other meteorological parameters.
Severe heat events are influenced by a complex interplay of meteorological and socio-economic factors. In addition to temperature, parameters such as geopotential height, precipitation, vegetation index, and land surface albedo contribute to event severity. To assess the potential impact of heatwaves, population distribution and vulnerability data are integrated into the analysis. By training models on historical heatwave events, incorporating meteorological parameters and mortality counts, we have developed a predictive capability for heat-related mortality. While our initial models exhibit some overestimation, they accurately capture overall trends. Ongoing research focuses on refining input data and model structure to enhance prediction accuracy.
We have developed an automated monitoring system that routinely processes new satellite data to detect anomalies and severe events in near real-time. Information is distributed through an interactive website, enabling timely responses and informed decision-making. Our efforts in real-time detection of severe events facilitate proactive responses, build resilient systems, and enhance overall preparedness for unforeseen events.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner