'Weather Forecasting Is Simply Not A Luxury!': Deborah Ross Goes Nuclear On Trump Over NOAA Cuts

Forbes Breaking News

During a House Science, Space and Tech Committee hearing in July, Rep. Deborah Ross (D-NC) spoke about President Trump's cuts to the National Oceanic and Atmospheric Administration.

Transcript

00:00I now recognize my colleague from North Carolina, Ms. Ross, for five minutes.

00:04Thank you, Chairman Franklin and Ranking Member Ammo, for holding this very important and timely hearing.

00:11And thank you to our witnesses for joining us today.

00:14As we witnessed devastating floods in Texas, New Mexico, and my home state of North Carolina,

00:22displacing families, damaging infrastructure, and costing so many lives,

00:27we're reminded that weather forecasting isn't just science, it is survival.

00:34Emerging technologies are transforming how we collect and interpret weather data, and that is a good thing.

00:42From AI-powered models and high-resolution satellites to drone-based atmospheric sensing,

00:49we can now equip our outstanding forecasters at the National Weather Service

00:54with improved tools to predict weather patterns with increasing precision.

00:59And that's happening in the private sector as well.

01:02When these technologies are combined with data from traditional sources like weather balloons and radar systems,

01:09we gain a clearer, more comprehensive view of weather patterns than ever before.

01:15However, as we've heard, this progress is under threat.

01:20The Trump administration has proposed sweeping cuts to our nation's weather agencies,

01:26including the closure of all federally funded meteorology labs.

01:32Eliminating them means communities like those in Texas, New Mexico, and North Carolina,

01:38and, of course, across the country, will face the next disaster with less warning and fewer tools to respond.

01:47Consider the breakthrough by students at Stanford University who developed a cutting-edge AI forecasting model

01:54that rivals and, in some cases, surpasses existing systems.

01:59That model depends heavily on high-quality, long-term government weather data,

02:06data that has now been discontinued due to budget cuts.

02:11Emerging technologies hold immense promise, but they can't function without funding.

02:18And if we do not have sustained investment in core data and scientific research,

02:25even the most innovative solutions will not produce the warnings that we need.

02:32Weather forecasting is simply not a luxury.

02:34It's a critical service that protects communities, informs farmers, supports transportation, and saves lives.

02:43We must fight in a bipartisan way to preserve and strengthen our weather agencies and not dismantle them.

02:52Artificial intelligence is reshaping the future of weather modeling,

02:57offering new opportunities to generate faster, more localized, and potentially more accurate forecasts.

03:03And thank you, Dr. Gupta, for sharing with us about this.

03:07These advances are being driven by improvements in computing,

03:11access to large-scale environmental data sets, innovation, and foundational models.

03:15As the technology evolves, it raises important questions about how NOAA can harness its potential

03:22to improve public forecasting and resilience.

03:26Dr. Gupta and then Mr. Cavett.

03:29Because AI-based models rely on historical weather data to determine their output,

03:35areas with limited previously recorded data,

03:39and we heard a little bit about this with the Texas farms, right,

03:42particularly in rural and low-income communities, may be more challenging to detect.

03:48What are the significant challenges with data limitation?

03:51How can we best unlock innovation in weather predicting systems,

03:56particularly given their potential impacts on emergency preparedness and public safety decisions

04:01in these areas where you cannot get the information?

04:05So, Dr. Gupta first.

04:06Thank you for the question.

04:10I would say this is not completely true at the moment with large-scale machine learning models

04:15that can actually combine data from satellites, global-scale data,

04:19with highly localized data and generalized locations

04:22which may not have the same sensors available.

04:26So these are the technologies that actually allow you to combine data

04:29from many different kinds of sensors

04:31that are not always easy to use with existing physical models.

04:36Like, if you talk to people on the NOAA side, people will tell you,

04:39like, even with respect to existing observations,

04:42they're probably not able to use vast majority of them in their current operational models.

04:48So from that perspective, AI is actually very interesting and transformative

04:52in terms of using even existing data much better.

04:57Mr. Cavett?

04:58First, I'll say, you know, AI as an innovation,

05:01it's opening many doors in this industry.

05:04It allows us to ingest more data.

05:06It reduces the computational costs of these models.

05:09But it also requires training data.

05:12And for us to provide better data in these communities,

05:17we need higher resolution data in those areas in order to train the models to be effective.

05:22I would posit that one of the core ways we do that is from space.

05:25We need higher revisit observations from space,

05:28similar to what the commercial industry is building,

05:30in order to enable those models to really deliver the value that they can.

05:35And I yield that.

05:40Thanks for listening.

Category

Transcript

Be the first to comment

Recommended