Future weather forecasting will likely have more features


The international scientific society SPIE held its annual meeting, Aug. 11-15, at the San Diego Convention Center, and an Aug. 14 plenary session addressed the future of sensors in weather forecasting.

Sid Boukabara, who is the principal scientist for the Center for Satellite Applications and Research at the National Oceanic and Atmospheric Administration, gave a presentation titled “Thoughts on the Future of NOAA’s Satellite Remote Sensing for Weather Forecasting and Environment Monitoring.” Boukabara focused on driving factors for sensor-based weather forecasting and on the challenges and opportunities.

“The remote sensing of the future is going to be an integrated system,” Boukabara said.

The observation systems will include air-based platforms, satellite systems and ground-based or surface platforms. Boukabara clarified that the designs are envisioned for the late 2020-2030 timeframe so the actual architecture may differ.

“Satellite remote sensing is going to play a bigger and bigger role,” Boukabara said.

The users of the weather forecasting data will include agriculture, disaster preparedness agencies, insurance companies, fisheries and energy providers.

“From a societal point of view remote sensing has become more and more important,” Boukabara said. “There is an increased demand for accurate global satellite data.”

The processing includes calibration, transformation, aggregation, validation and quality monitoring.

“The process here is to go from raw measurements,” Boukabara said.

Measurements involve several factors.

“That signal is a mixture of a number of things in the earth environment,” Boukabara said.

The specific information being sought determines where on the spectrum the sensor is used.

“There is no sensor that can measure everything,” Boukabara said.

Drones and private space flight will aid in increased air-based and space-based observation platforms.

“Access to space has increased dramatically,” Boukabara said. “That will lead to new opportunities, new sensors, new satellites.”

Technological advances also create the potential problem of spectrum interference, especially with fifth-generation wireless communications technology.

“That is in conflict in some areas with satellite remote sensing,” Boukabara said.

The purchase and deployment of remote sensing infrastructure must take the market into account.

“The satellites that we are interested in need to be tailored toward the users’ needs,” Boukabara said.

More specific information and forecasting for smaller geographic areas are among the expectations of improved sensing for weather forecasts.

“The trend is toward more capabilities and higher resolution,” Boukabara said. “That will have an impact on the need to have more satellite data.”

Data fusion will provide opportunities for added value, although the data must be filtered into relevant information.

“Intelligent data fusion will likely become a major focus of remote sensing,” Boukabara said.

Currently approximately 7% of recorded data are selected as suitable for use and approximately 3% of the data are actually used.

“This data solution has some serious challenges now,” Boukabara said.

Another issue is that the increased special resolution will require additional power consumption, and Boukabara predicts that by 2025 the data will exceed the affordable power limit.

“Significant increase in satellite data volume is going to force us into new approaches,” Boukabara said.

The issue of what to deploy will likely involve coordination by the Department of Defense, including NASA, other nations’ governmental agencies and commercial interests.

“Decision making is to happen in the future,” Boukabara said.

The first polar-based wind vector wind lidar satellite was launched by the European Space Agency in August 2018.

“The data from that sensor are promising and are showing some good results,” Boukabara said.

Certain information currently can’t be measured from space.

“Surface pressure is a big gap,” Boukabara said.

Boukabara expected optimal configurations to be in operation during the 2030-2040 timeframe.

“The remote sensing of the future is going to have to adapt,” he said. “Remote sensors will have to adapt to be really agile.”

Joe Naiman can be reached by email at jnaiman@reedermedia.com.