Curs to NOAA US National Weather Service

Indeed, but the data it’s based on doesn’t drastically change every few years so it’s an entirely different concept.
 
lustyd said:
Not at all. Here we’re talking about LLMs it’s totally different than the AI used for weather.
The problem that you’re not seeing is that it’s very hard to introduce new knowledge in a statistical LLM so the “most informed scientists” create a view and that view /stays present and is impossible to eradicate. In the old days the status quo died with old experts as a famous quote said, if you want to introduce new ideas teach the young and the old ideas literally die off. You can’t teach old dogs new tricks easily. AI LLMs ironically make learning more accessible but make new ideas almost impossible so will harm science over all unless used properly. Even when used properly, a single good new idea is still drowned out by reams of accepted doctrine.

This is the same issue as experts refusing to stay abreast of new ideas, but it’s reinforced because the machine confirms they’re “right”. The only way to make an LLM provide new information is to first find it and ask the LLM to summarise it in a response.
Click to expand...
You are so wrong. Climate scientists do not create a view. They study what’s is happening and draw conclusions. They will often use statistical tests to rule out some processes and identify other approaches. They keep aspects of a subject under review until such time as there is no or minimal doubt. There is now no doubt that burning fossil fuels is creating climate change. There is doubt about how to tackle the problem. That is what is happening now and is involving a multidisciplinary approach.

As with weather (as opposed to climate) forecasting, new ideas will always be considered seriously. Anyone who considers that they have new ideas or a new approach is free to submit ideas or proposals to their national weather service and the IPCC. Experts do not refuse to consider new ideas or new data. That has been so with AI.


By the 1980s, researchers started exploring the use of AI to improve the accuracy of weather predictions. Data driven models were used to help prediction of weather for satellite launches. Availability of resources limited AI research while NWP models were still advancing. The recent increase in attention to AI/ML is a result of that realisation.
 
Ignoring what I say as usual. I can never tell if you’re trying to learn of desperate not to. I don’t have the energy to drag you to knowledge a second time so will step away.
 
The deleting of the data is an absolute scandal akin to book burning in my opinion. Staff can be rehired once sanity returns but data, and especially large data sets may never be recovered. Some agencies the staff took backups or sent copies overseas to other agencies but I suspect these are too large to even consider that approach.
 
I do not know which is of most concern. There is still a lot of clear, unequivocal information there about climate change and the reasons. However, when I did a little searching some pages were shown as not available. Of course, climate research and other activities will continue around the world and of high quality but, inevitably in smaller quantities.

Cut backs to day to day operations are a concern. All the US based views I have seen to concentrate on national forecasting and warning services. Money could be saved there using more automation and commercial services although all the observing network, radar and space based systems would still need to be maintained. What they seem to overlook is the international dimension. Weather services worldwide depend on international cooperation and coordination of effort.

Observations over the USA are just as important as all other data. Their forecast model is widely used although a little inferior to other models. The satellite component is essential although not now unique. At present, meteorology would suffer greatly, although probably not catastrophically if NOAA and the NWS went tits up.
 
Scientists attending the recent European Geosciences Union meeting in Vienna have said how cuts to science funding and the dismantling of climate agencies in the US is impacting their work. This is a slightly edited summary of some comments. They are chilling.


1. The decision to cut funding on climate science in the US will likely have cascading effects on both European research and global climate resilience. The transatlantic partnership is built on shared values and decades of cooperation. It has been crucial for advancing global understanding of climate change through joint efforts in Earth observation, including data infrastructure and joint policy innovation.

For example, several European Centre for Medium-Range Weather Forecasts products that rely on National Oceanic and Atmospheric Administration (NOAA) data streams will now face gaps in input data.

A major loss has been the defunding of Famine Early Warning Systems Network – a critical system that operated in more than 30 famine-prone countries and was the only consistent source of regular famine and food insecurity early warning.


These actions will slow our ability to understand and respond effectively to the climate crisis, a loss that, ultimately, impacts everyone.


2. I’m concerned about the credibility of science and communication. Attacks on science and scientific integrity are being driven by social media and alternative media platforms – and artificial intelligence (AI) systems can generate something that looks like a scientific paper, [but] is complete rubbish. [AI] basically drives down the cost of fakes to zero. We are running a heightened risk of climate misinformation.

Large social media networks and Grok [an AI system] can produce something like a paper in no time. Trying to undermine [science] by flooding the zone with sh*t is a declared strategy – and it is the total antithesis to science.

It is one thing to rebut people that are paid to ask stupid questions about the validity of climate science, as we’ve done for decades. [But it is another] to fight an AI system that produces all this bullsh*t in no time. As a scientific community, we are not equipped to fight back…I hope other actors in society are thinking about how to support climate science and the integrity of science and its role in society.


3. The NOAA cuts are definitely impacting the air-quality research community and rippling to Europe. For several months this year, our NOAA colleagues were not allowed to join online meetings with us anymore, stalling the fruitful discussions and collaborations we have had for many years. One NOAA colleague was supposed to give an invited talk at EGU, but he was not allowed to travel.

Thinking back to last year’s EGU where we hung out with many NOAA colleagues, it was such a short time ago – but seems almost like another era where trans-Atlantic scientific cooperation and US leadership in atmospheric research were a given. Now our US colleagues are scared – scared to put anything in writing that could be used against them, scared of losing their funding and/or jobs.


4. The current attacks on climate research in the US are extremely concerning. A full generation of young researchers have been fired from their positions and are unlikely to stay in climate research. While some might move to Europe or other continents, it is not in the interest of the global climate research community to see such a weakening and even a destruction of climate research in the US.

While some decisions against climate research institutions and researchers might be annulled by the courts, several might be near impossible to reverse. For instance, buildings [that] will no longer be available for research institutions and researchers [that] move on to other positions out of concerns for their job security.
​
​
​
Finally, Attacks on climate science won’t change the facts and the reality of climate change – but they weaken our possibilities to address the climate crisis.​
 
Last edited:
You must log in or register to reply here.

Share this page

Top