-
Diffusion models for probabilistic precipitation generation from atmospheric variables
Authors:
Michael Aich,
Sebastian Bathiany,
Philipp Hess,
Yu Huang,
Niklas Boers
Abstract:
Improving the representation of precipitation in Earth system models (ESMs) is critical for assessing the impacts of climate change and especially of extreme events like floods and droughts. In existing ESMs, precipitation is not resolved explicitly, but represented by parameterizations. These typically rely on resolving approximated but computationally expensive column-based physics, not accounti…
▽ More
Improving the representation of precipitation in Earth system models (ESMs) is critical for assessing the impacts of climate change and especially of extreme events like floods and droughts. In existing ESMs, precipitation is not resolved explicitly, but represented by parameterizations. These typically rely on resolving approximated but computationally expensive column-based physics, not accounting for interactions between locations. They struggle to capture fine-scale precipitation processes and introduce significant biases. We present a novel approach, based on generative machine learning, which integrates a conditional diffusion model with a UNet architecture to generate accurate, high-resolution (0.25°) global daily precipitation fields from a small set of prognostic atmospheric variables. Unlike traditional parameterizations, our framework efficiently produces ensemble predictions, capturing uncertainties in precipitation, and does not require fine-tuning by hand. We train our model on the ERA5 reanalysis and present a method that allows us to apply it to arbitrary ESM data, enabling fast generation of probabilistic forecasts and climate scenarios. By leveraging interactions between global prognostic variables, our approach provides an alternative parameterization scheme that mitigates biases present in the ESM precipitation while maintaining consistency with its large-scale (annual) trends. This work demonstrates that complex precipitation patterns can be learned directly from large-scale atmospheric variables, offering a computationally efficient alternative to conventional schemes.
△ Less
Submitted 31 March, 2025;
originally announced April 2025.
-
Conditional diffusion models for downscaling & bias correction of Earth system model precipitation
Authors:
Michael Aich,
Philipp Hess,
Baoxiang Pan,
Sebastian Bathiany,
Yu Huang,
Niklas Boers
Abstract:
Climate change exacerbates extreme weather events like heavy rainfall and flooding. As these events cause severe losses of property and lives, accurate high-resolution simulation of precipitation is imperative. However, existing Earth System Models (ESMs) struggle with resolving small-scale dynamics and suffer from biases, especially for extreme events. Traditional statistical bias correction and…
▽ More
Climate change exacerbates extreme weather events like heavy rainfall and flooding. As these events cause severe losses of property and lives, accurate high-resolution simulation of precipitation is imperative. However, existing Earth System Models (ESMs) struggle with resolving small-scale dynamics and suffer from biases, especially for extreme events. Traditional statistical bias correction and downscaling methods fall short in improving spatial structure, while recent deep learning methods lack controllability over the output and suffer from unstable training. Here, we propose a novel machine learning framework for simultaneous bias correction and downscaling. We train a generative diffusion model in a supervised way purely on observational data. We map observational and ESM data to a shared embedding space, where both are unbiased towards each other and train a conditional diffusion model to reverse the mapping. Our method can be used to correct any ESM field, as the training is independent of the ESM. Our approach ensures statistical fidelity, preserves large-scale spatial patterns and outperforms existing methods especially regarding extreme events and small-scale spatial features that are crucial for impact assessments.
△ Less
Submitted 5 April, 2024;
originally announced April 2024.
-
Fast, Scale-Adaptive, and Uncertainty-Aware Downscaling of Earth System Model Fields with Generative Machine Learning
Authors:
Philipp Hess,
Michael Aich,
Baoxiang Pan,
Niklas Boers
Abstract:
Accurate and high-resolution Earth system model (ESM) simulations are essential to assess the ecological and socio-economic impacts of anthropogenic climate change, but are computationally too expensive to be run at sufficiently high spatial resolution. Recent machine learning approaches have shown promising results in downscaling ESM simulations, outperforming state-of-the-art statistical approac…
▽ More
Accurate and high-resolution Earth system model (ESM) simulations are essential to assess the ecological and socio-economic impacts of anthropogenic climate change, but are computationally too expensive to be run at sufficiently high spatial resolution. Recent machine learning approaches have shown promising results in downscaling ESM simulations, outperforming state-of-the-art statistical approaches. However, existing methods require computationally costly retraining for each ESM and extrapolate poorly to climates unseen during training. We address these shortcomings by learning a consistency model (CM) that efficiently and accurately downscales arbitrary ESM simulations without retraining in a zero-shot manner. Our approach yields probabilistic downscaled fields at a resolution only limited by the observational reference data. We show that the CM outperforms state-of-the-art diffusion models at a fraction of computational cost while maintaining high controllability on the downscaling task. Further, our method generalizes to climate states unseen during training without explicitly formulated physical constraints.
△ Less
Submitted 26 February, 2025; v1 submitted 5 March, 2024;
originally announced March 2024.
-
Design and implementation of a noise temperature measurement system for the Hydrogen Intensity and Real-time Analysis eXperiment (HIRAX)
Authors:
Emily R. Kuhn,
Benjamin R. B. Saliwanchik,
Maile Harris,
Moumita Aich,
Kevin Bandura,
Tzu-Ching Chang,
H. Cynthia Chiang,
Devin Crichton,
Aaron Ewall-Wice,
Austin A. Gumba,
N. Gupta,
Kabelo Calvin Kesebonye,
Jean-Paul Kneib,
Martin Kunz,
Kavilan Moodley,
Laura B. Newburgh,
Viraj Nistane,
Warren Naidoo,
Deniz Ölçek,
Jeffrey B. Peterson,
Alexandre Refregier,
Jonathan L. Sievers,
Corrie Ungerer,
Alireza Vafaei Sadr,
Jacques van Dyk
, et al. (2 additional authors not shown)
Abstract:
This paper describes the design, implementation, and verification of a test-bed for determining the noise temperature of radio antennas operating between 400-800MHz. The requirements for this test-bed were driven by the HIRAX experiment, which uses antennas with embedded amplification, making system noise characterization difficult in the laboratory. The test-bed consists of two large cylindrical…
▽ More
This paper describes the design, implementation, and verification of a test-bed for determining the noise temperature of radio antennas operating between 400-800MHz. The requirements for this test-bed were driven by the HIRAX experiment, which uses antennas with embedded amplification, making system noise characterization difficult in the laboratory. The test-bed consists of two large cylindrical cavities, each containing radio-frequency (RF) absorber held at different temperatures (300K and 77K), allowing a measurement of system noise temperature through the well-known 'Y-factor' method. The apparatus has been constructed at Yale, and over the course of the past year has undergone detailed verification measurements. To date, three preliminary noise temperature measurement sets have been conducted using the system, putting us on track to make the first noise temperature measurements of the HIRAX feed and perform the first analysis of feed repeatability.
△ Less
Submitted 15 January, 2021;
originally announced January 2021.
