Accurate representations of locally meaningful future extreme weather events and the implications for India

It is now evident that periods of extreme temperature and humidity can transform buildings from places of shelter to sources of significant morbidity and mortality. Mitigating this risk requires computer representations of such events. Unfortunately, neither an agreed globally consistent representational method nor long-term continuous hourly weather data at a sufficient spatial resolution exist, from which such events can be extracted. Here, we introduce a new, mathematically sound, physically meaningful and consistent method to represent hot-dry, hot-wet and cold-dry extreme weather events of arbitrary length localised to the weather of the location in question, unlike existing methods that use absolute temperature thresholds regardless of locale. Importantly for human survival, this new method includes humidity. Replicable globally, we apply this method to India using carefully calibrated computer-generated weather data localised to 25 km (4790 locations) for now and the future, making India the first Global South country to obtain such an extreme event dataset. A series of tests show high consistency and reliability including low mean bias error against all known long-term (∼100 year) dry-bulb hot extremes of +0.2°C (σ = 2.4°C) and a mean deviation of 4.5°C (σ = 4.9°C) compared to the equivalent extreme period in a typical year. By moving a validated computer model of a typical apartment across all 4790 locations, we find significant, spatially varied, differences in indoor mean-maximum temperature and discomfort degree hours for a range of events, past (1981-2010) and future (2060-2089). Periods where mortality is likely to be significant are found. The reliability of the data, combined with their unprecedented spatiotemporal resolution and timescale, transforms our ability to study a wide range of weather-influenced phenomena, such as indoor and outdoor risks to human health or crop yields, under current and climate-changed weather. Practical Application: There is an urgent need for the public and professionals to fully understand that the impacts of climate change will be most transformative with respect to extreme, not mean, temperatures; with mass mortality becoming common. The method introduced here for the creation of extreme timeseries is general and replicable globally. The files produced for the case of India are freely available and already in use via a public repository. They are helpful to building designers, urban planners, policymakers and professionals in a variety of fields to analyse performance under extreme conditions and the implications for human health.