SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2005.10374 · 2020

Stochastic Super-Resolution for Downscaling Time-Evolving Atmospheric Fields with a Generative Adversarial Network

arXiv · PDF · Open in the Atlas

Code that ran

8 of 10 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 8 have no contract verdict; 0 failures of our instrument; 2 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2005.10374.json · badge: /api/ran/2005.10374.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
jleinonen/downscaling-rnn-gan canonical (classifier label)
classifier basis: census
8 ran · 2 no recorded run
FunctionStatusWhere it lives
MultiScaleSSIM Ran
ran on a synthesized input
jleinonen/downscaling-rnn-gan/dsrnngan/msssim.py
code served (permissive licence) · get_code("693f7625441b5824")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family float, shape (2,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create two small 4D image batches with the same shape
# Shape: [batch_size=2, height=64, width=64, channels=3]
batch_size = 2
height = 64
width = 64
channels = 3

np.random.seed(0)
img1 = np.random.randint(0, 256, size=(batch_size, height, width, channels), dtype=np.uint8)
img2 = np.random.randint(0, 256, size=(batch_size, height, width, channels), dtype=np.uint8)

# Define the arguments dictionary
ARGS = {
    "img1": img1,
    "img2": img2,
    "max_val": 255,
    "filter_size": 11,
    "filter_sigma": 1.5,
    "k1": 0.01,
    "k2": 0.03,
    "weights": None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (2,) — shape (2,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
balanced_spatial_average Ran
ran on a synthesized input
jleinonen/downscaling-rnn-gan/dsrnngan/rainfarm.py
code served (permissive licence) · get_code("8ec6dff5b9b4b4be")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 2, dtype family float, shape (4, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
from scipy.ndimage import convolve

# Create a small input array
x = np.random.randn(4, 5).astype(np.float32)

# Create a small kernel for convolution
# Using a 3x3 kernel
k = np.array([
    [0.1, 0.2, 0.1],
    [0.2, 0.4, 0.2],
    [0.1, 0.2, 0.1]
], dtype=np.float32)

ARGS = {
    "x": x,
    "k": k
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (4, 5) — shape (4, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
crps_ensemble Ran
ran on a synthesized input
jleinonen/downscaling-rnn-gan/dsrnngan/crps.py
code served (permissive licence) · get_code("17f1e068c67986de")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family float, shape (2,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Construct valid arguments for crps_ensemble
# The function expects:
# - observation: array-like of shape (...,) representing observed values
# - forecasts: array-like of shape (..., num_ensemble_members) representing ensemble forecasts

# Create a small observation array (batch of 2 observations)
observation = np.array([1.5, 2.3])

# Create a small forecasts array (batch of 2, with 4 ensemble members each)
forecasts = np.array([
    [0.8, 1.2, 1.6, 2.1],  # 4 forecasts for first observation
    [1.9, 2.1, 2.4, 2.8]   # 4 forecasts for second observation
])

ARGS = {
    "observation": observation,
    "forecasts": forecasts
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (2,) — shape (2,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
ensure_list Ran
ran on a synthesized input
jleinonen/downscaling-rnn-gan/dsrnngan/meta.py
code served (permissive licence) · get_code("0397bd2d5a36751a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

ARGS = {
    "x": "test_value"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
generator_initialized Ran
ran on a synthesized input
jleinonen/downscaling-rnn-gan/dsrnngan/models.py
code served (permissive licence) · get_code("2abd6399586eaf5b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import tensorflow as tf
from tensorflow.keras.models import Model
from tensorflow.keras.layers import Input, Layer, Concatenate
import numpy as np


# Create a minimal init_model that takes [lores_in_t0, noise_in_initial]
# and returns initial_state
class InitModel(Layer):
    def __init__(self, state_dim=64):
        super().__init__()
        self.state_dim = state_dim
    
    def call(self, inputs):
        # inputs is [lores_t0, noise_initial]
        # lores_t0 shape: (batch, H, W, num_channels)
        # noise_initial shape: (batch, H_noise, W_noise, 8)
        # Return a state tensor
        return tf.zeros((tf.shape(inputs[0])[0], 4, 4, self.state_dim))


# Create a minimal gen model that takes [lores_in, initial_state, noise_in_update]
# and returns [img_out, h]
class GenModel(Layer):
    def __init__(self, output_channels=1):
        super().__init__()
        self.output_channels = output_channels
    
    def call(self, inputs):
        # inputs is [lores_in, initial_state, noise_in_update]
        # lores_in shape: (batch, num_timesteps, H, W, num_channels)
        # initial_state shape: (batch, state_h, state_w, state_dim)
        # noise_in_update shape: (batch, num_timesteps, H_noise, W_noise, 8)
        batch_size = tf.shape(inputs[0])[0]
        # Generate output image
        img_out = tf.zeros((batch_size, 64, 64, self.output_channels))
        # Generate hidden state
        h = tf.zeros((batch_size, 4, 4, 64))
        return [img_out, h]


# Instantiate the models
init_model = InitModel(state_dim=64)
gen = GenModel(output_channels=1)

ARGS = {
    'gen': gen,
    'init_model': init_model,
    'num_channels': 1,
    'num_timesteps': 8,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
input_shapes Ran
ran on a synthesized input
jleinonen/downscaling-rnn-gan/dsrnngan/meta.py
code served (permissive licence) · get_code("469ba9d8738ccef8")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

class MockInput:
    def __init__(self, name, shape):
        self.name = name
        self.shape = shape

class MockModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.inputs = [
            MockInput("model_input_0", torch.Size([2, 4, 3])),
            MockInput("model_input_1", torch.Size([2, 5, 6])),
            MockInput("other_input_0", torch.Size([2, 7, 8])),
            MockInput("model_input_2", torch.Size([2, 9])),
        ]

model = MockModel()
prefix = "model_input"

ARGS = {
    "model": model,
    "prefix": prefix,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
log_slope Ran
ran on a synthesized input
jleinonen/downscaling-rnn-gan/dsrnngan/rainfarm.py
code served (permissive licence) · get_code("1d9fc5224f2c9ec9")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float64, ndim 0, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create sample data for log_k and log_power_spectrum
# These should be 1D numpy arrays with numeric values
np.random.seed(0)

log_k = np.linspace(0, 10, 50)  # 50 log wavenumber values
log_power_spectrum = 2 * log_k + np.random.normal(0, 0.5, 50)  # Linear relationship with noise

ARGS = {
    "log_k": log_k,
    "log_power_spectrum": log_power_spectrum,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
wasserstein_loss Ran
ran on a synthesized input
jleinonen/downscaling-rnn-gan/dsrnngan/gan.py
code served (permissive licence) · get_code("41421d55d26f881d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned EagerTensor, ndim 1, dtype family float, shape (2,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import tensorflow as tf
import numpy as np

# Create small test tensors for y_true and y_pred
# The function uses K.mean(y_true * y_pred, axis=-1)
# So we need compatible tensors that can be element-wise multiplied
# batch_size=2, feature_dim=4

y_true = tf.constant([[1.0, 0.0, 1.0, 0.0],
                      [0.0, 1.0, 0.0, 1.0]], dtype=tf.float32)

y_pred = tf.constant([[0.8, 0.2, 0.9, 0.1],
                      [0.3, 0.7, 0.2, 0.8]], dtype=tf.float32)

ARGS = {
    'y_true': y_true,
    'y_pred': y_pred
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 1, dtype family float, shape (2,) — shape (2,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
radial_average Not yet run
unverified (harvested, no recorded run)
jleinonen/downscaling-rnn-gan/dsrnngan/rainfarm.py
code served (permissive licence) · get_code("8effcdaf0cd7e987")
How it ran

no invocation recorded.

resize_lanczos Not yet run
unverified (harvested, no recorded run)
jleinonen/downscaling-rnn-gan/dsrnngan/plots.py
code served (permissive licence) · get_code("e0e94f7ff0dd7bcb")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 10 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Generative adversarial networks (GANs) have been recently adopted for super-resolution, an application closely related to what is referred to as "downscaling" in the atmospheric sciences: improving the spatial resolution of low-resolution images. The ability of conditional GANs to generate an ensemble of solutions for a given input lends itself naturally to stochastic downscaling, but the stochastic nature of GANs is not usually considered in super-resolution applications. Here, we introduce a recurrent, stochastic super-resolution GAN that can generate ensembles of time-evolving high-resolution atmospheric fields for an input consisting of a low-resolution sequence of images of the same field. We test the GAN using two datasets, one consisting of radar-measured precipitation from Switzerland, the other of cloud optical thickness derived from the Geostationary Earth Observing Satellite 16 (GOES-16). We find that the GAN can generate realistic, temporally consistent super-resolution sequences for both datasets. The statistical properties of the generated ensemble are analyzed using rank statistics, a method adapted from ensemble weather forecasting; these analyses indicate that the GAN produces close to the correct amount of variability in its outputs. As the GAN generator is fully convolutional, it can be applied after training to input images larger than the images used to train it. It is also able to generate time series much longer than the training sequences, as demonstrated by applying the generator to a three-month dataset of the precipitation radar data. The source code to our GAN is available at https://github.com/jleinonen/downscaling-rnn-gan.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2005.10374")
get_code_for_paper("2005.10374")
have("2005.10374")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2005.10374.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2005.10374.svg)](https://syntology.ai/paper/2005.10374)

Connect an agent — have() is free.