Published May 22, 2023
| Version
v1
Dataset
Open
Data for the manuscript "DEUCE v1.0: A neural network for probabilistic precipitation nowcasting with aleatoric and epistemic uncertainties" by Harnist et al. (2023)
Description
File descriptions:
verif_inputs.zip and verif_case_inputs.zip
Contain folders of raw PGM composites, containing needed input data for verification experiments, that is for the creation of baseline nowcast as well as observation HDF5 files.
deuce_inputs.zip
Contains HDF5 archives of the composite crops used for the training of DEUCE, derived from raw PGM composites using the create_hdf5_dataset.py script given.
deuce_model_checkpoints.py
Contains the two intermediate pytorch lightning checkpoints of DEUCE, as the final model checkpoint used for making predictions.
metrics.zip
Contains the netcdf files of verification metric values computed from DEUCE and baseline model predictions compared to observations.
Files
deuce_inputs.zip
Files
(7.5 GB)
| Name | Size | Download selected | Download all |
|---|---|---|---|
|
Checksum: md5:3a09d08f1c3b0e2dc53e93dffc778e8e
PID: http://hdl.handle.net/11304/1a133f85-75d1-4393-9cb8-79b0e3b3662b |
3.7 kB |
|
Download |
|
Checksum: md5:c0e7921fe1a75d394834bea73a213690
PID: http://hdl.handle.net/11304/d0c5c743-6c3d-4ef3-9ab0-016121ff7621 |
3.3 GB |
|
Preview Download |
|
Checksum: md5:1620ee2b9f88cd5736c3c80f94247a45
PID: http://hdl.handle.net/11304/415b944c-de34-4ae2-98b4-2ab672d66c38 |
2.7 GB |
|
Preview Download |
|
Checksum: md5:bfa2f33ce838f341673a2f284afafd4a
PID: http://hdl.handle.net/11304/f0466b68-2133-4a10-8593-d4a6bf97396d |
4.4 MB |
|
Preview Download |
|
Checksum: md5:8ca45e350fe78f3545505c1014e0d5dc
PID: http://hdl.handle.net/11304/51eee3a2-27db-4733-8f2d-3368c49d74f3 |
41.5 MB |
|
Preview Download |
|
Checksum: md5:90231d298330c1ec9b9fee1c12a367cf
PID: http://hdl.handle.net/11304/e3488d2e-e7f8-4213-8996-06001e4e8bbb |
1.5 GB |
|
Preview Download |
Additional details
Identifiers
- URL
- https://etsin.fairdata.fi/dataset/6b62077d-f0f4-4c24-a859-137c32c4df40
- B2SHARE Legacy Record ID
- 3efcfc9080fe4871bd756c45373e7c11
FMI metadata
- Process step
- Arrays of deuce_inputs.zip HDF5 files are derived from the original composites using create_hdf5_dataset.py. All other files are produced using the source code accompanying the work.
- Source data
- 11 polarimetric C-band radars in the FMI radar network
- Lineage
- The radar composites of verif_inputs.zip and verif_case_inputs.zip are original versions produced by FMI and no further quality control has been performed. Other work is derived from such composites.
- Parameter
- Data levels (meter, hectoPascal, degree, sigma pressure levels, other) in vertical direction(+/-) for example 1500 m or 850 hPa
- Resolution
- 1000
- Topic category
- climatologyMeteorologyAtmosphere
Parameter name
:
Radar reflectivity
Parameter unit
:
dBZ
Parameter description
:
Radar reflectivity compressed with offset=-32 and scale=0.5 for PGM archives, float32 values for DEUCE input HDF5 files.
Level
:
500 m
Temporal Coverage
Ranges:
Start date: 2019-04-30
End date: 2019-09-30
End date: 2019-09-30
Start date: 2020-04-30
End date: 2020-09-30
End date: 2020-09-30
Start date: 2021-04-30
End date: 2021-09-30
End date: 2021-09-30
Start date: 2022-04-30
End date: 2022-09-30
End date: 2022-09-30