Heatwaves arrive like clockwork over many regions of the world and are getting even more attention due to the upward trends in many regions in terms of their duration, frequency, intensity, and scale.
They arrive at more or less expected times of the year because they are directly related to the natural seasonal arrival of spring. This pattern is called the climate. Most people see heat waves on the horizon as March rolls in over the Indian subcontinent. This is why the old adage says climate is what you expect and the weather is what you get; and herein lies the challenge of accurately predicting heatwaves at a hyperlocal scale to save lives. Fortunately, India is getting better at predicting extremes and managing the resulting disasters.
Predicting a heatwave
The challenge of predicting heatwaves, or in fact any extremes, raises the question of how we should depict them for the people at large. For better or for worse, a new trend has emerged where even reputable scientific organisations, such as NASA, publicise frightening total-temperature maps. While these maps may communicate alarm well, they can also be misleading. Scientific methodology says we need to show anomalies, not the absolute value.
In reality, heat waves are excess temperatures over the expected seasonal values. Seasonally warm temperatures are expected and they don’t become heatwaves until we have anomalous temperatures riding on top of them. This is a simple point, yet it requires strict adherence to avoid confusing the general public. To illustrate, here are maps that compare visual representations of total and anomalous temperature values.
Credit: Akash Verma, IITB
The total temperature can be the average value over several days or the average of the daily minimum and daily maximum temperatures. Temperatures averaged over several days show the warmth (which we expect) over the ocean and on land, with lower temperatures over the Himalayas and the Tibetan Plateau. The daily minimum temperature map shows colder temperatures everywhere compared to the daily averages, again as one would expect. The daily maximum temperatures are similar to those in NASA’s maps — and they now appear routinely in media reports. And for good measure, some reports also use words like “frying pan” and “cauldron” to describe temperatures over large areas.
Note that even if specific countries are identified with their maximum temperatures, these annotations are not so useful for real-world disaster management. Total temperature maps hide many details of what we need from forecasts to guide disaster management. This fact becomes clear in the anomaly temperature maps for the same days. The contrasts with the total temperature maps couldn’t be starker than when we examine the anomalous quantities in the same temperatures. We derive these anomalies by subtracting the long-term average temperatures for the same days from the recorded values. For example, if the long-term average is 30 degrees C and the recorded value is 34 degrees C, the anomalous value is 4 degrees C.
The daily averages and minimum and maximum temperatures for the days considered here were in fact lower than normal over parts of India and Pakistan. Anomalies in the maximum temperature are indeed much lower than anomalies in the minimum temperatures over these regions. The warmest maximum temperatures occurred over several places — but this hardly describes a “frying pan” or a “cauldron”. However, Pakistan’s heatwave season extends into July, so it isn’t out of the woods yet as far as heatwaves this year are concerned.
Global warming is local
Another crucial aspect of mapping anomalies instead of total temperatures is that the location-specific processes that generate these heatwaves are very important for us to be able to predict them at the scales relevant to disaster management. Entire countries can’t prepare for weather disasters; the sharper the spatial hazard and the risk information, the likelier it is for disaster management to be effective and efficient. This is rendered more important considering the limited resources — even in rich countries, leave alone the economically developing world facing worse heatwaves.
Such caveats apply to extremes of rainfall, droughts, and floods as well. However, anomalous temperatures manifest on much larger scales. Exposing people to scary maps may raise awareness but they could also undermine credibility if a heat wave doesn’t materialise in a particular local area or if forecasts don’t pan out.
The case for anomaly maps
Forecasting systems function at the scale of the forecast’s model grid. For example, the Indian forecast system for short- (days 1-3) and medium- (days 3-10) range forecasts has a resolution of 12 km, so the forecasts are made at this scale. Early warnings are also developed based on these forecasts.
The National Disaster Management Agency coordinates with the India Meteorological Department and local governments to spread awareness across the country and save lives. With more than 12 lakh polling booths this year for the Lok Sabha polls, scary maps may also have the negative consequence of scaring people away from turning out to vote. We can avoid this by using anomaly maps instead. Reliable early warning systems can also be improved continuously, especially by downscaling the model forecasts to neighbourhood scales using artificial intelligence and machine-learning techniques. More examples of such successful downscaling are now being reported with applications in measuring and forecasting extreme events as well as in other sectors such as agriculture, energy, and health. Forecasts are also best downscaled for anomalies.
The writer is visiting professor, IIT Bombay and emeritus professor, University of Maryland.
