SoylentNews is people
SoylentNews
In May 2019, Neil Jacobs, the acting head of the National Oceanic and Atmospheric Administration (NOAA), testified before Capitol Hill that 5G wireless signals could decrease forecasting accuracy by 30 percent.
"This would degrade the forecast skill by up to 30%. If you look back in time to see when our forecast skill was roughly 30% less than it was today, it's somewhere around 1980," Jacobs said in May. "This would result in the reduction of hurricane track[ing] forecasts' lead time by roughly two to three days." A delay of two to three days could have a catastrophic effect on human life.
Still, these warnings haven't swayed regulators nor the cell phone industry. In August, Sprint announced more cities would be added to its 5G rollout plan. AT&T already has 5G available to corporate customers in various cities. Verizon already offers 5G to customers and has plans to expand, too.
"Right now the uncertainty is to what extent there will be an interference," he added. "In some sense the cause for education is to make sure that the existing weather sensing bands are protected and that 5G is in areas that are far enough away from where present weather sensors exist."
This does not mean 5G can't exist in states like Florida, but that the power might have to be turned down.
"If the power is turned down, there is a lesser likelihood that water satellites (that will sense the atmosphere) will sense the 5G network" instead, Gerth said.
Federal Communications Commission Chairman Ajit Pai has been dismissive of these concerns, which are only one of several in regards to 5G. As several experts told Salon last year, the effects of widespread use of mobile 5G need to be better-studied before it goes mainstream.
Why study when you can profit instead?
(Score: 3, Informative) by Anonymous Coward on Friday September 06 2019, @10:33PM
Your comment is bullshit and belongs at -1. I hope the moderators will mod you down where you belong.
Forecast skill has improved substantially over time, greatly reducing the errors at longer lead times. I'll just point you to NHC's statistics [noaa.gov] for track errors. Forecast tracks are based heavily on the output of computer models, so the substantial improvement in track errors is very much related to the skill of the forecast models. Forecasts were significantly worse 30 years ago.
The atmosphere is a chaotic system, meaning that small differences in the initial state of the system will grow exponentially over time. Models either represent the atmosphere on a grid or as a series of waves, as is the case with spectral models. Features smaller than the grid spacing or the smallest wave will not be resolved by the models. Some of the equations that the models solve, such as the Navier-Stokes equations, have no analytical solution. The models solve them numerically by integrating over discrete time steps. Increasing the resolution of the model requires smaller time steps, otherwise CFL errors [wikipedia.org] arise and the model becomes numerically unstable. The observations going into the models are also imperfect, just due to instrument error. Particularly in the upper atmosphere, where in-situ observations are very sparse, there's limited data to update the initial conditions of the model. It's actually pretty remarkable that computer models have as much skill as they do.
There are lots of scientists working continuously to improve the forecast models. These models do get thrown out and replaced with improved versions and, sometimes, entirely new models. Back on June 12, the old GFS model was scrapped in favor of the entirely new FV3 dynamical core. The new model is still called the GFS, but it's essentially an entirely new model, with a fundamentally different core. Your statement is just false.
Observation errors do have significant impacts on longer term forecasts because those small errors grow over time. Like I said, the atmosphere is a chaotic system. Some of the models are intentionally wrong to estimate the impact of errors in the initial state of the model on the resulting forecast. The same model is run many times with slightly different initial conditions to see how the forecast changes. This is how most ensemble forecasts work, and that's where the spaghetti plots of forecast tracks come from.
Hurricanes are steered by winds in the mid-levels of the troposphere, so upper air observations are needed to accurately predict forecast tracks. Aside from satellites, there just aren't a lot of these observations over the ocean. Even over land, radiosonde launches are usually every 12 hours and are much sparser than surface observations. Water vapor imagery from satellites is used to estimate the mid- and upper-level winds in the atmosphere. These observations are very important, not just for measuring atmospheric moisture, but also for the winds. Particularly over water, but also over land, satellite observations of moisture and the derived winds are very important to the models. Radiosonde observations are just too sparse.
You might respond and say that 5G observations won't have much of an impact on satellite observations over the oceans. However, weather systems over land do influence the track of tropical cyclones. The position of a trough over the eastern US might be the difference between a hurricane recurving and staying out over the Atlantic versus making a landfall along the east coast. Because the atmosphere is a chaotic system, it's very important to minimize the errors in the initial state of the models.
In short, you haven't got a clue what you're talking about. And forecasts were much worse 30 years ago. The NHC track errors prove this.