3 ms·

Since the headline is positively wrong, and the article paywalled here's what this actually is about: NOAA uses cloud providers including AWS and GCP to host a

Since the headline is positively wrong, and the article paywalled here's what this actually is about: NOAA uses cloud providers including AWS and GCP to host a bunch of services and sites under its "research division". The contracts under which these hosting arrangements are provided are to be terminated shortly. At that point those services and web sites will go offline. afaics it's nothing to do with the actual weather service data feeds.

financetechbro 2y ago
Read the excerpt I posted in this thread from Project 2025. Hope that gives you a clear perspective of the plan Edit: typo
rimunroe 2y ago
> the headline is positively wrong The headline is "US Weather Agency Websites Set to Vanish with Contract Cuts". Are you seeing a different version than I am? Your complaints don't make sense otherwise. > NOAA uses cloud providers including AWS and GCP to host a bunch of services and sites under its "research division". The contracts under which these hosting arrangements are provided are to be terminated shortly. At that point those services and web sites will go offline. The headline says that sites are going to go down because of contracts being cut. That seems like an accurate summary of this paragraph. > it's nothing to do with the actual weather service data feeds The headline doesn't say anything about data feeds.
floatrock 2y ago
My job requires pulling large data dumps produced by research from a national lab from public S3 buckets. TB's of data. This research is publicly funded, the data produced is public data, and it enables an entire ecosystem of business activity. imho, this is a good use of public funding and research, and hosting it in S3 buckets is a good solution that makes huge quantities of data easily accessible to people who would use this. I don't use NOAA data, but this has chilling effects for my business (is my national lab next?). There is nothing "efficient" about forcing agencies to get rid of data and service hosting that works well at scale. (If DOGE was actually about increasing efficiency of digital services like it's pre-rename/pre-takeover US Digital Services agency was actually doing, we'd be building a TB or PB-scale S3 alternative for government agencies to use, but we all know that's not the constructive efficiency they're doing here)
sunshine-o 2y ago
Thank you for the summary. Funny I was looking for some data from NOAA a few months ago and I ended up giving up. Their website has become a mess. I remember doing the same a few decades ago it was pretty straightforward, if I remember correctly there was a well organised anonymous FTP server and you would connect and transfer what you needed. So why not seeing this as an opportunity to go back to a simple, cheap and effective system? how hard can it be to serve those data with HTTP on a web server?
- floatrock 2y ago
  > how hard can it be to serve those data with HTTP on a web server? We're talking about TB and TB of scientific data, so pretty hard. Meanwhile, here's a nice well-organized easily-accessible glorified FTP server of all this: https://aws.amazon.com/marketplace/pp/prodview-mmz4ve4u5cdui#resources https://aws.amazon.com/marketplace/pp/prodview-mmz4ve4u5cdui... https://registry.opendata.aws/noaa-nexrad/ https://registry.opendata.aws/noaa-nexrad/
- alabastervlog 2y ago
  https://www.wpc.ncep.noaa.gov/grib/hpcgrib.shtml https://www.wpc.ncep.noaa.gov/grib/hpcgrib.shtml > The WPC makes available its Day 1-3 6-hourly QPF files in GRIB 1 format at 32km resolution. The GRIB files are available for download via the NCEP anonymous FTP server (ftp.hpc.ncep.noaa.gov) Et c., there's more. I think that's forecast data. Also: https://psl.noaa.gov/data/gridded/ https://psl.noaa.gov/data/gridded/ And so on. Throw "grib" (a data format they use a ton) on a search and you'll start finding pages like this. Took about a minute to find this stuff.

financetechbro 2y ago

Read the excerpt I posted in this thread from Project 2025. Hope that gives you a clear perspective of the plan Edit: typo

rimunroe 2y ago

> the headline is positively wrong The headline is "US Weather Agency Websites Set to Vanish with Contract Cuts". Are you seeing a different version than I am? Your complaints don't make sense otherwise. > NOAA uses cloud providers including AWS and GCP to host a bunch of services and sites under its "research division". The contracts under which these hosting arrangements are provided are to be terminated shortly. At that point those services and web sites will go offline. The headline says that sites are going to go down because of contracts being cut. That seems like an accurate summary of this paragraph. > it's nothing to do with the actual weather service data feeds The headline doesn't say anything about data feeds.

floatrock 2y ago

My job requires pulling large data dumps produced by research from a national lab from public S3 buckets. TB's of data. This research is publicly funded, the data produced is public data, and it enables an entire ecosystem of business activity. imho, this is a good use of public funding and research, and hosting it in S3 buckets is a good solution that makes huge quantities of data easily accessible to people who would use this. I don't use NOAA data, but this has chilling effects for my business (is my national lab next?). There is nothing "efficient" about forcing agencies to get rid of data and service hosting that works well at scale. (If DOGE was actually about increasing efficiency of digital services like it's pre-rename/pre-takeover US Digital Services agency was actually doing, we'd be building a TB or PB-scale S3 alternative for government agencies to use, but we all know that's not the constructive efficiency they're doing here)

sunshine-o 2y ago

Thank you for the summary. Funny I was looking for some data from NOAA a few months ago and I ended up giving up. Their website has become a mess. I remember doing the same a few decades ago it was pretty straightforward, if I remember correctly there was a well organised anonymous FTP server and you would connect and transfer what you needed. So why not seeing this as an opportunity to go back to a simple, cheap and effective system? how hard can it be to serve those data with HTTP on a web server?

floatrock 2y ago
> how hard can it be to serve those data with HTTP on a web server? We're talking about TB and TB of scientific data, so pretty hard. Meanwhile, here's a nice well-organized easily-accessible glorified FTP server of all this: https://aws.amazon.com/marketplace/pp/prodview-mmz4ve4u5cdui#resources https://aws.amazon.com/marketplace/pp/prodview-mmz4ve4u5cdui... https://registry.opendata.aws/noaa-nexrad/ https://registry.opendata.aws/noaa-nexrad/
alabastervlog 2y ago
https://www.wpc.ncep.noaa.gov/grib/hpcgrib.shtml https://www.wpc.ncep.noaa.gov/grib/hpcgrib.shtml > The WPC makes available its Day 1-3 6-hourly QPF files in GRIB 1 format at 32km resolution. The GRIB files are available for download via the NCEP anonymous FTP server (ftp.hpc.ncep.noaa.gov) Et c., there's more. I think that's forecast data. Also: https://psl.noaa.gov/data/gridded/ https://psl.noaa.gov/data/gridded/ And so on. Throw "grib" (a data format they use a ton) on a search and you'll start finding pages like this. Took about a minute to find this stuff.