Skip to content

Data access

Johannes Tobiassen Langvatn edited this page Sep 18, 2020 · 46 revisions

This wiki is being changed very often, any missing information or unclear information is welcome as suggestions in the issues-page

How to access latest weather forecast data from met.no

The weather forecast model output is available as NetCDF files on https://thredds.met.no/thredds/metno.html and the files follow the NetCDF/CF conventions. This page also documents how to navigate and access the data archived here. This page also documents how to access subsets of these files (So as to not have to download multiple GB of data)

Note also, that all communication to the thredds-server at Met requires a HTTPS connection, any HTTP-request reply with http error code 410: (410 Gone).

Navigating Thredds

When accesing the Weather Forecasts directory, the following screen greets you:

image.png

For how to access the file itself see Subsetting

MetCoOp Ensemble Prediction System (MEPS)

This is the "main" weather forecasting product. The current domain covers Norway, Sweden, Finland, and Estonia. The model is run with 75 seconds timesteps for a total of 66 hours, with output saved for each hour. It has a horizontal grid of 2.5km. The current version is run as a 30-member ensemble valid each 6th hour (TODO: Describe this, and the timeshiftiness)

See MEPS description for more details and a domain map (TODO).

MET AROME Arctic

This is an arctic weather forecasting product. The current domain covers Northern Norway, Svalbard and other arctic islands (See domain map in the AA description). The model is run with 75 seconds timesteps for a total of 66 hours, with output saved for each hour. It has a horizontal grid of 2.5km. It is currently only run as a single member per 6. hour, but this is likely to change

See AA description for more details and a domain map (TODO).

MET post-processed products

Described somewhat in Post-processed products

Other products

This includes ocean-modeldata and others. Will be explained more (Currently explained on: https://drive.google.com/file/d/0B3vvF5TQcILySGZEX1ViY2VlYms/view)

Old forecast products

The title says it all: these are products that were produced as part of the daily operation but now kept for archiving. Will be explained more (Currently explained on: https://drive.google.com/file/d/0B-SaEtrDE91WWDkzY3dpZlU1V00/view)

Subsetting

Following are some examples of how to access the files from the thredds-server using the OPeNDAP-protocol. There is also an option to download using a HTTPS-Server, though this is not adviced due to the file sizes involved

Subsetting using OPeNDAP web access form

Subsetting using NetCDF libraries

See Python, guides will be added for R and (maybe) a lower level (C-type) language

Subsetting using Fimex

How to read GRIB2 files using Fimex

The WMO has approved the GRIB-format for efficiently transmitting meteorological gridded data, so this is one of the preferred formats for exchanging data between meteorological services. GRIB has been around for some 35 years - since before the Internett and the PC was invented - with minimal changes to the format.

Being able to read GRIB data yourself, gives you the capacity to access gridded meteorological information from the weather services on the lowest possible level. This can be an advantage if you are looking for a reliable starting point for your data processing chain, which you plan to run with as little interruption as possible for the next decades to come.

Some basic receipts for installing and running the Fimex software for reading and interpolation GRIB2 files on Ubuntu are readily available here.

Clone this wiki locally