TORONTO - The mass grounding of commercial aircraft fleets during the Covid-19 pandemic substantially reduced the accuracy of weather forecasts around the world. The decreased flights during the spring eliminated about 75% of aircraft observations putting weather forecasts in peril.
Weather forecasts play an essential part in planning daily life, agricultural and economic activity. Meteorological observations collected from aircrafts are vital to accurate weather forecasting.
A new study, published in a journal by the American Geophysical Union, examines global weather forecasts 1-8 days ahead of occurrence. It has been comparing those forecasts against the best estimates of atmospheric conditions – and their impact – in a Covid-19 pandemic world for “relative accuracy”.
Ying Chen, researcher, and lead author in the study, compared forecasts with global temperatures, wind and precipitation data. There has been a shortage of such data recently.
Very briefly, here is how it works. During a flight, commercial airplanes log a variety of meteorological data including air temperature, relative humidity, air pressure and wind direction. This information is used to generate weather prediction models.
The lack of such crucial aircraft data, due to grounded planes, impact weather forecasting and impose additional economic costs, adding to the damage the pandemic has inflicted on economic activity. In effect, airplanes serve as “mobile weather stations”.
Accurate weather forecasts benefit the agricultural and energy sectors, and aids in stabilizing the electrical grid. For instance, wind turbines use wind to generate electricity. Accurate forecasts of windspeed and temperature will help energy companies predict the energy load they can provide to consumers each day.
“If this uncertainty goes over a threshold, it will introduce unstable voltage for the electrical grid”, Chen said in a press release. “That could lead to a blackout, and I think this is the last thing we want to see in this pandemic”.
Findings in the study show a large deterioration of forecast accuracy across North America, southeast China, and Australia. These are regions where in air traffic is typically heavy, and data therefore more reliable…and necessary.
However, forecasts in remote regions where data is limited, such as, Greenland, Siberia, Antarctica, and the Sahara Desert has also suffered, even if the impact is not as significantly relevant – low population levels and limited economic activity. The study found that the accuracy of forecasting models applicable to these regions was substantially worse during March-May 2020 when compared to February 2020 and previous years (2017-2019).
The research also highlights the deterioration in long-term forecasts and the reliability for early warnings of extreme weather – hurricanes and the like.
A surprising find in the study was that the seeming consistency of the weather forecasts over Western Europe. Despite a drop of approximately 90% in air traffic during the height of the pandemic, weather forecasts in the region remained relatively accurate.
This is probably due to the fact that the region is more reliant on the high density of conventional ground-based observation posts enabling forecasters to populate models with sufficient meteorological data.
Replicating a dense network of other types of sensors in various locations around the globe could help minimize the impacts of data losses during a global emergency in the future. Whether it is economically feasible or efficient is another story.