An official website of the United States government
Abstract Atmospheric waves in the tropical tropopause layer are recognized as a significant influence on processes that impact global climate. For example, waves drive the quasi‐biennial oscillation (QBO) in equatorial stratospheric winds and modulate occurrences of cirrus clouds. However, the QBO in the lower stratosphere and thin cirrus have continued to elude accurate simulation in state‐of‐the‐art climate models and seasonal forecast systems. We use first‐of‐their‐kind profile measurements deployed beneath a long‐duration balloon to provide new insights into impacts of fine‐scale waves on equatorial cirrus clouds and the QBO just above the tropopause. Analysis of these balloon‐borne measurements reveals previously uncharacterized fine‐vertical‐scale waves (<1 km) with large horizontal extent (>1000 km) and multiday periods. These waves affect cirrus clouds and QBO winds in ways that could explain current climate model shortcomings in representing these stratospheric influences on climate. Accurately simulating these fine‐vertical‐scale processes thus has the potential to improve sub‐seasonal to near‐term climate prediction.
more »
« less
Regression Forest Approaches to Gravity Wave Parameterization for Climate Projection
Abstract We train random and boosted forests, two machine learning architectures based on regression trees, to emulate a physics‐based parameterization of atmospheric gravity wave momentum transport. We compare the forests to a neural network benchmark, evaluating both offline errors and online performance when coupled to an atmospheric model under the present day climate and in 800 and 1,200 ppm CO2global warming scenarios. Offline, the boosted forest exhibits similar skill to the neural network, while the random forest scores significantly lower. Both forest models couple stably to the atmospheric model, and control climate integrations with the boosted forest exhibit lower biases than those with the neural network. Integrations with all three data‐driven emulators successfully capture the Quasi‐Biennial Oscillation (QBO) and sudden stratospheric warmings, key modes of stratospheric variability, with the boosted forest more accurate than the random forest in replicating their statistics across our range of carbon dioxide perturbations. The boosted forest and neural network capture the sign of the QBO period response to increased CO2, though both struggle with the magnitude of this response under the more extreme 1,200 ppm scenario. To investigate the connection between performance in the control climate and the ability to generalize, we use techniques from interpretable machine learning to understand how the data‐driven methods use physical information. We leverage this understanding to develop a retraining procedure that improves the coupled performance of the boosted forest in the control climate and under the 800 ppm CO2scenario.
more »
« less
- Award ID(s):
- 2004572
- PAR ID:
- 10526664
- Publisher / Repository:
- DOI PREFIX: 10.1029
- Date Published:
- Journal Name:
- Journal of Advances in Modeling Earth Systems
- Volume:
- 16
- Issue:
- 7
- ISSN:
- 1942-2466
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract We present single‐column gravity wave parameterizations (GWPs) that use machine learning to emulate non‐orographic gravity wave (GW) drag and demonstrate their ability to generalize out‐of‐sample. A set of artificial neural networks (ANNs) are trained to emulate the momentum forcing from a conventional GWP in an idealized climate model, given only one view of the annual cycle and one phase of the Quasi‐Biennial Oscillation (QBO). We investigate the sensitivity of offline and online performance to the choice of input variables and complexity of the ANN. When coupled with the model, moderately complex ANNs accurately generate full cycles of the QBO. When the model is forced with enhanced CO2, its climate response with the ANN matches that generated with the physics‐based GWP. That ANNs can accurately emulate an existing scheme and generalize to new regimes given limited data suggests the potential for developing GWPs from observational estimates of GW momentum transport.more » « less
-
Abstract Rising atmospheric CO2levels place terrestrial ecosystems under novel environmental conditions, and research in field settings is key to understanding how real plant communities will respond. Despite decades of progress in elevated CO2(eCO2) experiments, major gaps persist in our knowledge of plant responses to interacting influences of climate change, especially in areas outside North America and Western Europe.With a goal to expand access to field‐based eCO2research, we designed, built, and tested TinyCO2, a low‐cost field experiment for climate change research on plants. TinyCO2features sixteen 0.62‐m2plot areas, half with ambient and half with elevated (+200 ppm) CO2concentrations, and is suitable for short‐stature plants (≤0.5 m in height).Using a proportional‐integral control algorithm and constant sampling of air within the plots, TinyCO2achieves consistent elevation of [CO2] averaging +196.9 ppm. During testing, 95.1% of measured CO2concentrations fell within 20% of the setpoint (ambient CO2 + 200 ppm). A streamlined design and efficient use of instrumentation reduced the cost of the system to roughly one‐fifth of the cost of similar experiments from the past 30 years ($13.68 vs. $64.65 ppm−1 m−2, adjusted to 2024 USD).Our results demonstrate a system capable of precise and accurate field‐based CO2elevation for significantly reduced cost. We envision the TinyCO2design being implemented in a multitude of field‐based eCO2studies, perhaps as part of a globally distributed collaborative network experiment.more » « less
-
The response of forests to climate change depends in part on whether the photosynthetic benefit from increased atmospheric CO 2 (∆C a = future minus historic CO 2 ) compensates for increased physiological stresses from higher temperature (∆T). We predicted the outcome of these competing responses by using optimization theory and a mechanistic model of tree water transport and photosynthesis. We simulated current and future productivity, stress, and mortality in mature monospecific stands with soil, species, and climate sampled from 20 continental US locations. We modeled stands with and without acclimation to ∆C a and ∆T, where acclimated forests adjusted leaf area, photosynthetic capacity, and stand density to maximize productivity while avoiding stress. Without acclimation, the ∆C a -driven boost in net primary productivity (NPP) was compromised by ∆T-driven stress and mortality associated with vascular failure. With acclimation, the ∆C a -driven boost in NPP and stand biomass (C storage) was accentuated for cooler futures but negated for warmer futures by a ∆T-driven reduction in NPP and biomass. Thus, hotter futures reduced forest biomass through either mortality or acclimation. Forest outcomes depended on whether projected climatic ∆C a /∆T ratios were above or below physiological thresholds that neutralized the negative impacts of warming. Critically, if forests do not acclimate, the ∆C a /∆T must be above ca . 89 ppm⋅°C −1 to avoid chronic stress, a threshold met by 55% of climate projections. If forests do acclimate, the ∆C a /∆T must rise above ca . 67 ppm⋅°C −1 for NPP and biomass to increase, a lower threshold met by 71% of projections.more » « less
-
The present study investigates dynamical coupling between the equatorial stratospheric Quasi31 biennial oscillation (QBO) and the boreal winter surface climate of the Northern Hemisphere mid and high latitudes using 42 years data (1979–2020). For neutral El Niño Southern Oscillation (ENSO) periods, QBO westerlies (W) at 70 hPa favor high sea level pressure in the polar region, colder conditions and deeper snow over Eurasia and North America, and the opposite effects for QBO easterlies (E). When QBO anomalies arrive in the upper troposphere and lower stratosphere (UTLS), it is observed that planetary wave activity is enhanced in the extratropical UTLS during QBO W and diminished during QBO E. This QBO teleconnection pathway along the UTLS to the high latitude surface is independent of the “stratospheric pathway” (Holton-Tan mechanism). Diagnosis of this pathway can help to improve understanding of internal sub-seasonal to seasonal variations, and long-range forecasting over Eurasia and North America.more » « less
