Facial Recognition Technology Could Now Predict and Improve Weather Forecasting

by August 21, 2019 0 comments

Facial Recognition

A recently published study from the National Center for Atmospheric Research (NCAR) revealed that machine learning typically used in facial recognition could help in predicting extreme weather events and their austerity. Storms that produce hail can have large, damaging impacts on agriculture, property, and even wildlife and responsible for billions of dollars of damage every year.

For this study, researchers have trained a deep learning model called Convolution Neural Network (CNN) to locate specific storm features that can determine the formation of hail, and how large the hailstones will be.

In a statement, NCAR scientist David John Gagne said, “We know that the structure of a storm affects whether the storm can produce hail. A supercell is more likely to produce hail than a squall line, for example. But most hail forecasting methods just look at a small slice of the storm and can’t distinguish the broader form and structure.”

When looking at the reports, around 13,000 shorebirds and waterfowl were killed in a severe hailstorm in eastern Montana just last week. Every year, such storm costs more than US$22 billion in damages to people and property. As per the NOAA’s Severe Storms database, over 4,600 major hailstorm events were recorded in 2018, with the majority of damages being reported in the central part of the US.

The current research was supported by the National Science Foundation, an NCAR’s sponsor. On this research, Nick Anderson, an NSF program officer, stated that “Hail – particularly large hail – can have significant economic impacts on agriculture and property. Using these deep learning tools in unique ways will provide additional insight into the conditions that favor large hail, improving model predictions. This is a creative, and very useful, merger of scientific disciplines.”

 

Training Machine Learning Model on Random Forecasts

For the new study, the NCAR scientists turned to a type of machine learning model designed to study visual images. They trained the model using images of simulated storms, along with information about temperature, pressure, wind speed, and direction as inputs and simulations of hail resulting from those conditions as outputs.

The weather simulations were created using the NCAR-based Weather Research and Forecasting model (WRF). Afterward, the model figured out which features of the storm are correlated with, whether or not, it hails and how big the hailstones are. The ML-based model was trained and it demonstrated storm features that the scientists had previously linked to hailstones.

 

The NCAR Overview

The National Center for Atmospheric Research (NCAR) is a US federally funded research and development center (FFRDC) managed by the non-profit University Corporation for Atmospheric Research (UCAR) and backed by the National Science Foundation (NSF). The research center is staffed by scientists, engineers, technicians, and support personnel, who work in key areas of Climate, Metrology or Weather, Pollution and air chemistry, Space weather, Environmental and societal impacts, among others.

No Comments so far

Jump into a conversation

No Comments Yet!

You can be the one to start a conversation.

Your data will be safe!Your e-mail address will not be published. Also other data will not be shared with third person.