leap-stc / data-management

Collection of code to manually populate the persistent cloud bucket with data
https://catalog.leap.columbia.edu/
Apache License 2.0
0 stars 5 forks source link

New Dataset [Reconstructed Precipitation and Temperature CFSv2 Forecasts for 2011-2018] #15

Closed AlexandreRebiere closed 8 months ago

AlexandreRebiere commented 1 year ago

Dataset Name

Reconstructed Precipitation and Temperature CFSv2 Forecasts for 2011-2018

Dataset URL

https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/CEFZLV

Description

Temperature and precipitation forecasts from operational CFSv2 (Climate Forecast System), for biweekly dates in 2011-2018, with a prediction window of 3-6 weeks and interpolated to a 1x1 grid containing the western contiguous United States. This dataset is used in the Rodeo forecasting model, it is the root of our future work consisting in adding new features to the model.

Size

The dataset consists in 65Mo files in a total of 12GB.

License

Unknown

Data Format

Other (describe below)

Data Format (other)

Mots of them are in the pickle library (.pkl), some are .h5

Access protocol

HTTP(S)

Source File Organization

There is two files per month starting in April 2011 and finishing in April 2018

Example URLs

https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/CEFZLV

Authorization

No; data are fully public

Transformation / Processing

There's nothing to modify in this dataset. The forecasting model is able to work with this dataset. No more analysis is needed.

Target Format

Other

Comments

they have to keep the same name and the same format (.pkl and .h5)

jbusecke commented 8 months ago

I assume this is not needed anymore?