skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Estimating Full Longwave and Shortwave Radiative Transfer with Neural Networks of Varying Complexity
Abstract Radiative transfer (RT) is a crucial but computationally expensive process in numerical weather/climate prediction. We develop neural networks (NN) to emulate a common RT parameterization called the Rapid Radiative Transfer Model (RRTM), with the goal of creating a faster parameterization for the Global Forecast System (GFS) v16. In previous work we emulated a highly simplified version of the shortwave RRTM only—excluding many predictor variables, driven by Rapid Refresh forecasts interpolated to a consistent height grid, using only 30 sites in the Northern Hemisphere. In this work we emulate the full shortwave and longwave RRTM—with all predictor variables, driven by GFSv16 forecasts on the native pressure–sigma grid, using data from around the globe. We experiment with NNs of widely varying complexity, including the U-net++ and U-net3+ architectures and deeply supervised training, designed to ensure realistic and accurate structure in gridded predictions. We evaluate the optimal shortwave NN and optimal longwave NN in great detail—as a function of geographic location, cloud regime, and other weather types. Both NNs produce extremely reliable heating rates and fluxes. The shortwave NN has an overall RMSE/MAE/bias of 0.14/0.08/−0.002 K day−1for heating rate and 6.3/4.3/−0.1 W m−2for net flux. Analogous numbers for the longwave NN are 0.22/0.12/−0.0006 K day−1and 1.07/0.76/+0.01 W m−2. Both NNs perform well in nearly all situations, and the shortwave (longwave) NN is 7510 (90) times faster than the RRTM. Both will soon be tested online in the GFSv16. Significance StatementRadiative transfer is an important process for weather and climate. Accurate radiative transfer models exist, such as the RRTM, but these models are computationally slow. We develop neural networks (NNs), a type of machine learning model that is often computationally fast after training, to mimic the RRTM. We wish to accelerate the RRTM by orders of magnitude without sacrificing much accuracy. We drive both the NNs and RRTM with data from the GFSv16, an operational weather model, using locations around the globe during all seasons. We show that the NNs are highly accurate and much faster than the RRTM, which suggests that the NNs could be used to solve radiative transfer inside the GFSv16.  more » « less
Award ID(s):
2019758
PAR ID:
10512817
Author(s) / Creator(s):
; ; ;
Publisher / Repository:
AMS Journals
Date Published:
Journal Name:
Journal of Atmospheric and Oceanic Technology
Volume:
40
Issue:
11
ISSN:
0739-0572
Page Range / eLocation ID:
1407 to 1432
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; (, Journal of Atmospheric and Oceanic Technology)
    Abstract This paper describes the development of U-net++ models, a type of neural network that performs deep learning, to emulate the shortwave Rapid Radiative-transfer Model (RRTM). The goal is to emulate the RRTM accurately in a small fraction of the computing time, creating a U-net++ that could be used as a parameterization in numerical weather prediction (NWP). Target variables are surface downwelling flux, top-of-atmosphere upwelling flux ( ), net flux, and a profile of radiative-heating rates. We have devised several ways to make the U-net++ models knowledge-guided, recently identified as a key priority in machine learning (ML) applications to the geosciences. We conduct two experiments to find the best U-net++ configurations. In Experiment 1, we train on non-tropical sites and test on tropical sites, to assess extreme spatial generalization. In Experiment 2, we train on sites from all regions and test on different sites from all regions, with the goal of creating the best possible model for use in NWP. The selected model from Experiment 1 shows impressive skill on the tropical testing sites, except four notable deficiencies: large bias and error for heating rate in the upper stratosphere, unreliable for profiles with single-layer liquid cloud, large heating-rate bias in the mid-troposphere for profiles with multi-layer liquid cloud, and negative bias at lowzenith angles for all flux components and tropospheric heating rates. The selected model from Experiment 2 corrects all but the first deficiency, and both models run ~10 4 times faster than the RRTM. Our code is available publicly. 
    more » « less
  2. ; ; ; ; ; ; ; ; (, Journal of Geophysical Research: Atmospheres)
    Abstract Clouds and radiation play an important role in warming events over the Southern Ocean (SO). Here we evaluate European Center for Medium‐Range Weather Forecasts Reanalysis version 5 (ERA5) and Polar Weather Research Forecast (PWRF) output through comparison to surface‐based measurements of clouds, radiation, and the atmospheric state over the SO during 2017–2023 at Escudero Station (62.2°S, 58.97°W) on King George Island. ERA5 mean monthly downward shortwave (DSW) radiative fluxes are found to be 38–50 W m−2higher than observations in summer, whereas ERA5 mean monthly downward longwave (DLW) is biased by −18 to −22 W m−2in summer and −16 W m−2on average over the year. Comparisons of temperature, humidity, and lowest‐cloud base heights between ERA5 and observations rule these factors out as large contributors to the DLW flux biases. The similarity between observed DLW cloud forcing distributions for atmospheric columns containing low‐level liquid and ice‐only clouds suggests limited influence of cloud phase errors on DLW biases. Thus the most likely explanation for DLW flux biases in ERA5 is underestimated cloud optical depth, which is also consistent with DSW flux biases. Similar biases in ERA5 are found during atmospheric river (AR) events. By contrast, PWRF flux bias magnitudes are much smaller during AR events (−12 W m−2for DSW and −2 W m−2for DLW). After bias correction, ERA5 monthly average net cloud forcing over 2017–2023 is found to be a minimum of −107 W m−2in January and a maximum of 65 W m−2in June. 
    more » « less
  3. ; ; ; ; ; (, Atmospheric Chemistry and Physics)
    Abstract. Recent analyses show the importance of methane shortwave absorption, which many climate models lack. In particular, Allen et al. (2023) used idealized climate model simulations to show that methane shortwave absorption mutes up to 30 % of the surface warming and 60 % of the precipitation increase associated with its longwave radiative effects. Here, we explicitly quantify the radiative and climate impacts due to shortwave absorption of the present-day methane perturbation. Our results corroborate the hypothesis that present-day methane shortwave absorption mutes the warming effects of longwave absorption. For example, the global mean cooling in response to the present-day methane shortwave absorption is -0.10±0.07 K, which offsets 28 % (7 %–55 %) of the surface warming associated with present-day methane longwave radiative effects. The precipitation increase associated with the longwave radiative effects of the present-day methane perturbation (0.012±0.006 mm d−1) is also muted by shortwave absorption but not significantly so (-0.008±0.009 mm d−1). The unique responses to methane shortwave absorption are related to its negative top-of-the-atmosphere effective radiative forcing but positive atmospheric heating and in part to methane's distinctive vertical atmospheric solar heating profile. We also find that the present-day methane shortwave radiative effects, relative to its longwave radiative effects, are about 5 times larger than those under idealized carbon dioxide perturbations. Additional analyses show consistent but non-significant differences between the longwave versus shortwave radiative effects for both methane and carbon dioxide, including a stronger (negative) climate feedback when shortwave radiative effects are included (particularly for methane). We conclude by reiterating that methane remains a potent greenhouse gas. 
    more » « less
  4. ; ; ; (, Atmospheric Chemistry and Physics)
    Abstract. Mineral aerosols (i.e., dust) can affect climate and weather by absorbing and scattering shortwave and longwave radiation in the Earth's atmosphere, the direct radiative effect. Yet understanding of the direct effect is so poor that the sign of the net direct effect at top of the atmosphere (TOA) is unconstrained, and thus it is unknown if dust cools or warms Earth's climate. Here we develop methods to estimate the instantaneous shortwave direct effect via observations of aerosols and radiation made over a 3-year period in a desert region of the southwestern US, obtaining a direct effect of -14±1 and -9±6 W m−2 at the surface and TOA, respectively. We also generate region-specific dust optical properties via a novel dataset of soil mineralogy from the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS), which are then used to model the dust direct radiative effect in the shortwave and longwave. Using this modeling method, we obtain an instantaneous shortwave direct effect of -21±7 and -1±7 W m−2. The discrepancy between the model and observational direct effect is due to stronger absorption in the model, which we interpret as an AVIRIS soil iron oxide content that is too large. Combining the shortwave observational direct effect with a modeled longwave TOA direct effect of 1±1 W m−2, we obtain an instantaneous TOA net effect of -8±6 W m−2, implying a cooling effect of dust. These findings provide a useful constraint on the dust direct effect in the southwestern United States. 
    more » « less
  5. ; ; ; ; ; ; ; ; (, Journal of Advances in Modeling Earth Systems)
    Abstract Neural networks (NNs) are increasingly used for data‐driven subgrid‐scale parameterizations in weather and climate models. While NNs are powerful tools for learning complex non‐linear relationships from data, there are several challenges in using them for parameterizations. Three of these challenges are (a) data imbalance related to learning rare, often large‐amplitude, samples; (b) uncertainty quantification (UQ) of the predictions to provide an accuracy indicator; and (c) generalization to other climates, for example, those with different radiative forcings. Here, we examine the performance of methods for addressing these challenges using NN‐based emulators of the Whole Atmosphere Community Climate Model (WACCM) physics‐based gravity wave (GW) parameterizations as a test case. WACCM has complex, state‐of‐the‐art parameterizations for orography‐, convection‐, and front‐driven GWs. Convection‐ and orography‐driven GWs have significant data imbalance due to the absence of convection or orography in most grid points. We address data imbalance using resampling and/or weighted loss functions, enabling the successful emulation of parameterizations for all three sources. We demonstrate that three UQ methods (Bayesian NNs, variational auto‐encoders, and dropouts) provide ensemble spreads that correspond to accuracy during testing, offering criteria for identifying when an NN gives inaccurate predictions. Finally, we show that the accuracy of these NNs decreases for a warmer climate (4 × CO2). However, their performance is significantly improved by applying transfer learning, for example, re‐training only one layer using ∼1% new data from the warmer climate. The findings of this study offer insights for developing reliable and generalizable data‐driven parameterizations for various processes, including (but not limited to) GWs. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Text Encoded Conversion
  • XML
  • Save / Share this Record
  • Send to Email