You are driving down the road and your smartphone or vehicle alert system announces that black ice is 1.5 miles ahead. As you move closer there’s a countdown: 500 feet, 300 feet, and now urgently: Warning, Warning — 100 feet to black ice.
How can you know the location of black ice so precisely?
Weather systems will collect data from other vehicles on that same road. The vehicles will wirelessly transmit road condition and weather data. Vehicles’ data points will include barometric pressure, air temperature, windshield wiper settings, and vehicle stability control, or the amount of differential rotation between wheels indicating slippery conditions. There will be data about the amount of sun and headlight status, among other metrics.
All this data will be collected by the sensors that underpin the Internet of Things: Temperature, pressure, moisture and light sensors as well as motion sensors such as accelerometers and gyroscopes. Many of these sensors are already in your cellphones, and will soon to be just about everywhere. This means weather measurements will also be everywhere, and this will improve the precision of weather condition reports and forecasting.
Before this sensor and wireless technology arrived, weather data primarily came from airport weather stations and ships at sea. Weather observers recorded conditions such as wind direction and speed, temperature, pressure and pressure trends, dew point and sky cover. Professionals plotted this data and then transmitted it for forecast analysis. In total, forecasters had hundreds of weather plots to work with, or if global forecasts were needed, thousands. The potential number of weather data points today, thanks to sensors, will be in the billions.
More data is giving rise to precision forecasting, which will not only be important to drivers, but to government as well, along with many industries, especially agriculture.
Two years ago, Schneider Electric began selling remote monitoring stations aimed at agriculture that measure atmospheric and ground conditions such as soil moisture. There are now 4,000 of these systems deployed.
That data, when combined with government and private sources of weather data, is used to help develop forecasts that update every hour and help farmers “make a better tactical decision” about when to apply pesticides, water and fertilizer, said Ron Sznaider, senior vice president of Schneider’s cloud service.
The idea “is to mitigate the risk of weather,” said Sznaider.
Weather systems in agriculture help farmers identify variations on their acreage that can change by elevation, hill or valley, or proximity to a water source. “That’s important to identifying micro-climates on farms,” said William Mahoney, deputy director of the applications research laboratory at the National Center for Atmospheric Research. By optimizing watering, for instance, farms may be able to save money in the long run, he said.
But utilizing weather data that is coming from an increasingly diverse array of sources is not so easy.
Weather researchers are now pushing for good metadata along with the sensor data. The metadata, or data about the data, will help researchers know which instruments were used to gather the data and their accuracy, so the people who run forecasting models “can pick and choose data sets that are going to be of value,” said Mahoney.
The sensor data that is now collected from vehicles is being used for research. The quality of this vehicle data can vary depending on where sensors are located, such as their proximity to the engine. Even the color of the car — light or dark — influences the accuracy of the sensor data. There are no standards yet for optimizing sensor placement.
There are potential life-saving benefits to using vehicle sensor data to improve weather information and alert systems for drivers. There are over 5.76 million car crashes a year, and approximately 22% of those crashes — or 1.3 million — are weather related, according to the Federal Highway Administration (FHA).
“On average, nearly 6,000 people are killed and over 445,000 people are injured in weather-related crashes each year,” reported the FHA.
There are plenty of challenges ahead. Increasing amounts of sensor data also means “trying to capture the physics correctly” at those finer resolutions, said Mahoney. The feedback from urban environments, lakes, rivers, streams and many other conditions all influence micro-climates. “Those physical interrelationships matter,” he said.
For now, weather forecasters aren’t routinely using cellphone or vehicle data “but that’s coming,” said Mahoney, who said that over the next few years, research will result in methods and techniques to take advantage of those data sources.
This story, “Internet of Things brings new era of weather forecasting” was originally published by
