skip to main content

Title: Using Neural Networks to Learn the Jet Stream Forced Response from Natural Variability

Two distinct features of anthropogenic climate change, warming in the tropical upper troposphere and warming at the Arctic surface, have competing effects on the midlatitude jet stream’s latitudinal position, often referred to as a “tug-of-war.” Studies that investigate the jet’s response to these thermal forcings show that it is sensitive to model type, season, initial atmospheric conditions, and the shape and magnitude of the forcing. Much of this past work focuses on studying a simulation’s response to external manipulation. In contrast, we explore the potential to train a convolutional neural network (CNN) on internal variability alone and then use it to examine possible nonlinear responses of the jet to tropospheric thermal forcing that more closely resemble anthropogenic climate change. Our approach leverages the idea behind the fluctuation–dissipation theorem, which relates the internal variability of a system to its forced response but so far has been only used to quantify linear responses. We train a CNN on data from a long control run of the CESM dry dynamical core and show that it is able to skillfully predict the nonlinear response of the jet to sustained external forcing. The trained CNN provides a quick method for exploring the jet stream sensitivity to a wide range of tropospheric temperature tendencies and, considering that this method can likely be applied to any model with a long control run, could be useful for early-stage experiment design.

more » « less
Award ID(s):
Author(s) / Creator(s):
 ;  ;  ;  
Publisher / Repository:
American Meteorological Society
Date Published:
Journal Name:
Artificial Intelligence for the Earth Systems
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Extratropical eddy-driven jets are predicted to shift poleward in a warmer climate. Recent studies have suggested that cloud radiative effects (CRE) may enhance the amplitude of such shifts. But there is still considerable uncertainty about the underlying mechanisms, whereby CRE govern the jet response to climate change. This study provides new insights into the role of CRE in the jet response to climate change by exploiting the output from six global warming simulations run with and without atmospheric CRE (ACRE). Consistent with previous studies, it is found that the magnitude of the jet shift under climate change is substantially increased in simulations run with ACRE. It is hypothesized that ACRE enhance the jet response to climate change by increasing the upper-tropospheric baroclinicity due to the radiative effects of rising high clouds. The lifting of the tropopause and high clouds in response to surface warming arises from the thermodynamic constraints placed on water vapor concentrations. Hence, the influence of ACRE on the jet shift in climate change simulations may be viewed as an additional “robust” thermodynamic constraint placed on climate change by the Clausius–Clapeyron relation. The hypothesis is tested in simulations run with an idealized dry GCM, in which the model is perturbed with a thermal forcing that resembles the ACRE response to surface warming. It is demonstrated that 1) the enhanced jet shifts found in climate change simulations run with ACRE are consistent with the atmospheric response to the radiative warming associated with rising high clouds, and 2) the amplitude of the jet shift scales linearly with the amplitude of the ACRE forcing.

    more » « less
  2. Uncertainty about the influence of anthropogenic radiative forcing on the position and strength of convective rainfall in the Intertropical Convergence Zone (ITCZ) inhibits our ability to project future tropical hydroclimate change in a warmer world. Paleoclimatic and modeling data inform on the timescales and mechanisms of ITCZ variability; yet a comprehensive, long-term perspective remains elusive. Here, we quantify the evolution of neotropical hydroclimate over the preindustrial past millennium (850 to 1850 CE) using a synthesis of 48 paleo-records, accounting for uncertainties in paleo-archive age models. We show that an interhemispheric pattern of precipitation antiphasing occurred on multicentury timescales in response to changes in natural radiative forcing. The conventionally defined “Little Ice Age” (1450 to 1850 CE) was marked by a clear shift toward wetter conditions in the southern neotropics and a less distinct and spatiotemporally complex transition toward drier conditions in the northern neotropics. This pattern of hydroclimatic change is consistent with results from climate model simulations indicating that a relative cooling of the Northern Hemisphere caused a southward shift in the thermal equator across the Atlantic basin and a southerly displacement of the ITCZ in the tropical Americas, with volcanic forcing as the principal driver. These findings are at odds with proxy-based reconstructions of ITCZ behavior in the western Pacific basin, where changes in ITCZ width and intensity, rather than mean position, appear to have driven hydroclimate transitions over the last millennium. This reinforces the idea that ITCZ responses to external forcing are region specific, complicating projections of the tropical precipitation response to global warming. 
    more » « less
  3. Abstract

    Separating the climate response to external forcing from internal climate variability is a key challenge. While most previous studies have focused on surface responses, here we examine zonal‐mean patterns of North Pacific subsurface temperature responses. In particular, the changes since 1950 driven by anthropogenic aerosol emissions are found by using a pattern recognition method. Based on the single‐forcing large‐ensemble simulations from two models, we show that aerosol forcing caused a nonmonotonic temporal response and a characteristic zonal‐mean pattern within North Pacific, which is distinct from the pattern associated with internal variability. The aerosol‐forced pattern with the nonmonotonic temporal feature shows a substantial temperature change in subpolar regions and a reversed change on the southern flank of the subtropical gyre. A similar characteristic pattern and nonmonotonic time evolution are extracted from the subsurface observations, which likely reflect the subsurface responses to the aerosol forcing, although differences exist with the simulated responses.

    more » « less
  4. Abstract

    The teleconnection between the Quasi‐Biennial Oscillation (QBO) and the boreal winter polar vortex, the Holton–Tan effect, is analyzed in the Whole Atmosphere Community Climate Model (WACCM) with a focus on how stationary wave propagation varies by QBO phase. These signals are difficult to isolate in reanalyses because of large internal variability in short observational records, especially when decomposing the data by QBO phase. A 1,500‐year ensemble is leveraged by defining the QBO index at five different isobars between 10 and 70 hPa. The Holton–Tan effect is a robust part of the atmospheric response to the QBO in WACCM with warming of the polar stratosphere during easterly QBO (QBOE). A nudging technique is used to reduce polar stratospheric variability in one simulation. This enables isolation of the impact of the QBO on the atmosphere in the absence of a polar stratospheric response to the QBO: referred to as the “direct effect” and the polar stratospheric response, “indirect effect.” This simulation reveals that the polar stratospheric warming during QBOE pushes the tropospheric jet equatorward, opposing the poleward shift of the jet by the QBOE, especially over the North Pacific. The Holton–Tan effect varies over longitude. The QBO induces stronger planetary wave forcing to the mean flow in the extratropical lower stratosphere between Indonesia and Alaska. The North Pacific polar stratosphere responds to this before other longitudes. What follows is a shift in the position of the polar vortex toward Eurasia (North America) during easterly (westerly) QBO. This initiates downstream planetary wave responses over North America, the North Atlantic, and Siberia. This spatiotemporal evolution is found in transient simulations in which QBO nudging is “switched on.” The North Pacific lower stratosphere seems more intrinsically linked to the QBO while other longitudes appear more dependent on the mutual interaction between the QBO and polar stratosphere.

    more » « less
  5. null (Ed.)
    Abstract. Changes in glacier length reflect the integrated response to local fluctuations in temperature and precipitation resulting from both external forcing (e.g., volcanic eruptions or anthropogenic CO2) and internal climate variability. In order to interpret the climate history reflected in the glacier moraine record, the influence of both sources of climate variability must therefore be considered. Here we study the last millennium of glacier-length variability across the globe using a simple dynamic glacier model, which we force with temperature and precipitation time series from a 13-member ensemble of simulations from a global climate model. The ensemble allows us to quantify the contributions to glacier-length variability from external forcing (given by the ensemble mean) and internal variability (given by the ensemble spread). Within this framework, we find that internal variability is the predominant source of length fluctuations for glaciers with a shorter response time (less than a few decades). However, for glaciers with longer response timescales (more than a few decades) external forcing has a greater influence than internal variability. We further find that external forcing also dominates when the response of glaciers from widely separated regions is averaged. Single-forcing simulations indicate that, for this climate model, most of the forced response over the last millennium, pre-anthropogenic warming, has been driven by global-scale temperature change associated with volcanic aerosols. 
    more » « less