license: artistic-2.0
task_categories:
- unconditional-image-generation
size_categories:
- 100K<n<1M
dataset_info:
- config_name: combined
features:
- name: image
dtype: image
- name: datetime
dtype: string
- name: ssn
dtype: float64
splits:
- name: train
num_bytes: 268736675
num_examples: 25000
download_size: 268932295
dataset_size: 268736675
- config_name: foe
features:
- name: image
dtype: image
- name: datetime
dtype: string
- name: ssn
dtype: float64
splits:
- name: train
num_bytes: 197419425
num_examples: 25000
download_size: 197597483
dataset_size: 197419425
- config_name: fof2
features:
- name: image
dtype: image
- name: datetime
dtype: string
- name: ssn
dtype: float64
splits:
- name: train
num_bytes: 261691250
num_examples: 25000
download_size: 262081847
dataset_size: 261691250
- config_name: hmf2
features:
- name: image
dtype: image
- name: datetime
dtype: string
- name: ssn
dtype: float64
splits:
- name: train
num_bytes: 288035775
num_examples: 25000
download_size: 287599117
dataset_size: 288035775
configs:
- config_name: combined
data_files:
- split: train
path: combined/train-*
default: true
- config_name: foe
data_files:
- split: train
path: foe/train-*
- config_name: fof2
data_files:
- split: train
path: fof2/train-*
- config_name: hmf2
data_files:
- split: train
path: hmf2/train-*
250k maps of ionospheric characteristics generated by IRI-2020 spanning random dates 1958-2024 and random SSN 0-200 (not the true SSN for that date), with the goal of training a diffusion inpainting model to extrapolate plausible ionospheres.
All maps are 1°x1° gridded, 361x181 with 180°W at the left, 180°E at the right, 90°N at the top, 90°S at the left. IRI settings are JF(4:6), JF(21:23), JF(28:30), JF(33:36), JF(39:40), JF(47) false, all other JF true.
Each file contains in PNG metadata Datetime: the date/time passed to IRI, as an ISO8601 string, and Ssn, the RZ(12) passed to IRI.
foe is FoE, scaled to 0 = 0.5 MHz, 255 = 5 MHz.fof2 is foF2, scaled to 0 = 1.5 MHz, 255 = 15 MHz.hmf2 is hmF2, scaled to 0 = 150 km, 255 = 450 km.mufd is MUF(3000km), scaled to 0 = 5 MHz, 255 = 45 MHz.
combined is an fof2 map in rows 0-180, a mufd map in rows 182-362, and an hmf2 map in rows 364-544. The intervening rows are cyan.