If you are a fan of the work we do each week with our Gestalt IT Rundown on Facebook, you probably saw a story in this week’s episode about the race for 5G spectrum causing some potential problems with weather forecasting. I didn’t have the time to dig into the details behind the story on that episode, so I wanted to take a few minutes and explain why it’s such a big deal.
First, you have to know that 5G (and many other) speeds are entirely dependent upon the amount of spectrum they can use to communicate. The more spectrum available to them, the more channels they have available to communicate. Which increases the speed they can exchange information and reduces the amount of interference between devices. Sounds simple right?
Except mobile devices aren’t the only things that are using the spectrum. We have all kinds of other devices out there that use radio waves to communicate. We’ve known for several years that there are a lot of devices in the 5 GHz spectrum used by 802.11 that interfere with wireless devices. Things like ISM radios for industrial and medical applications or government radar systems. The government has instituted many regulations in those frequency ranges to ensure that critical infrastructure isn’t interfered with.
When Nature Calls
However, sometimes you can’t regulate away interference. According to this Wired article the FCC, back in March, opened up auctions for the 24 GHz frequency band. This was over strenuous objections from NASA, NOAA, and the American Meteorological Society (AMS). Why is 24 GHz so important? Well, as it turns out, there’s a natural phenomenon that exists at that range.
Recall your kitchen microwave. How does it work? Basically, it uses microwave radiation to heat the water in the food you’re cooking. How does it do that? Turns out the natural frequency of water is 2.38 GHz. Now, thanks to the magic of math, 23.8 GHz is a multiple of that frequency. Which means that anything that broadcasts at 23.8 GHz will have issues with water, such as water in tree leaves or in water pipes.
So, why is NOAA and the AMS freaking out about auctioning off spectrum in the 23.8 GHz range? Because anything broadcasting in that range is not only going to have issues with water interference but it’s also going to look like water to sensitive equipment. That means that orbiting weather satellites that use microwaves to detect water vapor in the air that reacts to 23.8 GHz are going to encounter co-channel interference from 5G radio sources.
You might say to yourself, “So what? It’s just a little buzz, right?” Well, except that that little buzz creates interference in the data being fed into forecast prediction models. Those models are the basis for the weather forecasts we have today. And if you haven’t noticed the reliability of our long range forecasts has been steadily improving for the past 30 years or so. Today’s 7-day forecasts are almost 80% accurate, which is pretty good compared to how bad things were in the 80s, where you could only guarantee 80% accuracy from a 3-day forecast.
Guess what? NOAA says that if the 24 GHz spectrum gets auctioned off for 5G use, we could see the accuracy of our forecasting regress almost 30%, which would push our models back to where they were in the 80s. Now, for those of you that live in places that are fortunate enough to only get sun and the occasional rain shower that doesn’t sound too bad, right? Just make sure to pack an umbrella. But for those that live in places where there is a significant chance for severe weather, it’s a bit more problematic.
I live in Oklahoma. We’re right in the middle of Tornado Alley. In the spring between April 1 and June 1 my state becomes a fun place full of nasty weather that can destroy homes and cause widespread devastation. It’s never boring for sure. But in the last 30 years we’ve managed to go from being able to give people a few minutes warning about an impending tornado to being able to issue Potential Dangerous Situation (PDS) Tornado Watches up to 48 hours in advance. While a PDS Tornado Watch doesn’t mean that we’re going to get one in a specific area, it does mean that you need to be on the lookout for something that day. It gives enough warning to make sure you’re not going to get caught flat footed when things get nasty.
The easiest way to avoid this problem is probably the least likely to happen. The FCC needs to restrict the auction of that spectrum range identified by NOAA and NASA until it can be proven that there won’t be any interference or that the forecast accuracy isn’t going to be impacted. 5G rollouts are still far enough in the future that leaving a part of the spectrum out of the equation isn’t going to cause huge issues for the next few years. But if we have to start creating rules for how we have to change power settings for device manufacturers or create updates for fixed-position sensors and old satellites we’re going to have a lot more issues down the road than just slightly slow mobile devices.
The reason why this is hard is because an FCC focused on opening things up for corporations doesn’t care about the forecast accuracy of a farmer in Iowa. They care about money. They care about progress. And ultimately they care about looking good over saving lives. There’s no guarantee that reducing forecast accuracy will impact life saving, but the odds are that better forecasts will help people make better decisions. And ultimately, when you boil it down to the actual choices, the appearance is that the FCC is picking money over lives. And that’s a pretty easy choice for most people to make.
If I’m a bit passionate about weather tech, it’s because I live in one of the most weather-active places on the planet. The National Severe Storms Laboratory and the National Weather Center are both about 5-6 miles from my house. I see the use of this tech every day. And I know that it saves lives. It’s saved lives for years for people that need to know
There are opposing views due to evidence about the senors: https://www.ctia.org/news/how-a-fake-weather-sensor-could-take-out-5g . I suspect this might turn out to be more a political move than a technical one. It appears the interference testing by NASA and NOAA was inappropriate and the sensors in question do not exist. It appears your safety is not at risk due to this auction.
The CTIA is the wireless industry trade group, so it’s not surprising of their opinion. However, NASA and the DoD concur with NOAA and they are currently using sensors on the 23.8 GHz band: https://www.washingtonpost.com/weather/2019/05/23/head-noaa-says-g-deployment-could-set-weather-forecasts-back-years-wireless-industry-denies-it
It still did not address how the tests of interference were conducted. In other words, how the frequency is used, power and scope, for example. This was not addressed in this article referenced above at least and in the CTIA it does mention the newer sensors which are not as susceptible to the (I will use gov’t here to replace the different groups involved) govt’s standards of tests. But CTIA’s argument is that this is not how the frequency is used. In addition, the industry has been ramping this up for some time and the gov’t is just catching this now? In summary, the tests conducted by the gov’t were inaccurate and inapplicable to the way 5G would work.
Thinking about it, I would say that due to the nature of the design of the gov’t systems, we could be susceptible to outside influences from other unfriendly countries. Maybe the gov’t should fix that problem.
Part of the issue from my perspective is that the sensors are picking up a signal at 23.8 GHz because that’s the frequency that water resonates at. Any co-channel or adjacent frequency would cause interference in that band if not properly regulated. It’s the reason why UNI-2 and UNI-3 are highly regulated bands in the 5 GHz spectrum. Because any transmission in those bands could potentially interfere with the radar that detects in that range.
Perhaps the CTIA is being sincere in their assessment. But the linked piece seems to be less about an honest discussion of the merits of the issues and more about an agenda that’s being pushed by the 5G industry. Perhaps CTIA would consent to be interviewed by a third party to clear the record?