Date of Award
1-1-2023
Language
English
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
College/School/Department
Department of Atmospheric and Environmental Sciences
Content Description
1 online resource (x, 174 pages) : illustrations (some color)
Dissertation/Thesis Chair
Mathias Vuille
Committee Members
Brian Rose, Oliver Elison Timm
Keywords
Arctic, Atlantic, Pacific, precipitation, sea ice, South America, Atmospheric science, Environmental sciences, Climatic changes
Subject Categories
Atmospheric Sciences
Abstract
Atlantic and Pacific sea surface temperature (SST) variations mainly include El Niño–Southern Oscillation (ENSO) on interannual timescales and Pacific multidecadal variability (PMV) and Atlantic multidecadal variability (AMV) on decadal-interdecadal timescales. Understanding their impacts on South American and Arctic climate is crucial for studying climate variability and helpful to constrain climate model projections for these regions on multidecadal timescales. Prior studies focused mainly on their impacts on global climate, but multidecadal impacts on South American climate are poorly understood due to the lack of long observational records. Furthermore, earlier research investigated primarily the isolated impacts on global climate from the Atlantic or Pacific, but how they jointly affect South American precipitation and surface air temperature is not well documented.Arctic sea-ice concentration and thickness have been declining in recent decades largely due to greenhouse gas (GHG)-induced warming, but climate models forced with historical forcing underestimate recent Arctic sea-ice loss rates. This underestimation may be caused by multidecadal internal variability, such as the AMV and PMV. Therefore, understanding the AMV and PMV’s impacts on Arctic climate is critical. Here I first analyzed the AMV and PMV’s impacts on South American precipitation and temperature in observations and the associated anomalous atmospheric circulation in atmospheric reanalyses. The results show that South American precipitation and temperature anomaly patterns associated with the AMV and PMV depend on the phase of the other basin’s variability, and the precipitation and temperature anomalies are likely due to the changes in large-scale convection and subsidence associated with regional Hadley and Walker circulation anomalies over South America and the South Atlantic Convergence Zone (SACZ). During an AMV warm (cold) phase, PMV-induced July-June-August (JJA) precipitation anomalies are more positive (negative) over 0o-10oS and southeastern South America, but more negative (positive) over the northern Amazon and central Brazil. PMV-induced precipitation anomalies in December-January-February (DJF) are more positive (negative) over Northeast Brazil and southeastern South America during the warm (cold) AMV phase, but more negative (positive) over the central Amazon Basin and central-eastern Brazil. PMV’s impact on AMV-induced precipitation anomalies shows similar dipole patterns. In JJA, PMV- and AMV-induced temperature anomalies are more positive (negative) over all of South America when the other basin is in a warm (cold) phase, but in DJF temperature anomalies are more positive (negative) only over the central Andes and central-eastern Brazil and more negative (positive) over southeastern South America and Patagonia. Precipitation and temperature anomalies over central Brazil in JJA and southern Bolivia and northern Argentina in DJF are anti-correlated, suggesting that cloud cover and soil moisture feedbacks associated with the precipitation anomalies affect temperatures in these regions. Second, I compared South American precipitation and temperature anomalies associated with unsmoothed Atlantic and Pacific SST variations with those associated with the AMV and PMV. This analysis was performed using both observations and Community Atmospheric Model version 5 (CAM5) Atmospheric Model Intercomparison Project (AMIP) simulations. The model can reproduce the general precipitation and temperature anomalies seen in observations, such as warm-phase PMV-related drying over northern South America and wetting over northern Argentina during DJF, warm-phase AMV-related wetting over northern South America and drying over eastern Brazil during DJF; but the model also shows some discrepancies with observations with generally weaker signals, especially over the SACZ region. The results show that the Pacific SST variability’s influences on South American precipitation and temperature are different on the decadal and interannual timescales. ENSO and AMV signals dominate the unsmoothed Pacific and Atlantic SST variations, respectively. However, CAM5 cannot simulate the difference between the unsmoothed and smoothed Pacific SST variations’ impacts on South American precipitation and temperature. The Atlantic and Pacific SST variations modulate each other’s influences on South American rainfall in similar ways, and the modulation is generally similar in terms of pattern for interannual and multidecadal variability although its magnitude is stronger on the interannual timescale. The Atlantic and Pacific modulate each other’s influences on South American precipitation through changes in the horizontal moisture flux and vertical motion, which are responsible for precipitation changes over the SACZ and River Plate basin during DJF and equatorial regions during JJA, respectively. The results also show that, on multidecadal timescales, the AMV and PMV have the potential to modulate the anthropogenically-forced wetting trend over southeastern South America and the warming trend over all of South America, but their modulation effect is not large enough to account for the historical simulations’ underestimation of southeastern South American wetting trend. Third, I used CESM1 idealized and time-varying pacemaker ensemble simulations to investigate AMV and PMV’s contribution to Arctic sea ice changes in recent decades, and the underlying mechanisms. The pacemaker simulations nudge SST to observations only over the AMV or PMV regions while the rest of the world is fully coupled with observed radiative forcing in the atmosphere. These experiments show that recent PMV can contribute up to 0.32 1014 m2 (3.4% of the 1920-2013 mean) to Arctic sea-ice area variations and 0.31 1015 m3 (10.9%) to Arctic sea-ice volume variations during 1920-2013, while recent AMV may be responsible for 0.26 1014 m2 (2.8% of the 1920-2013 mean) and 0.32 1015 m3 (11.1%) of the Arctic sea-ice area and volume variations, respectively. The sea-ice concentration variability is mainly over the marginal Arctic Ocean, while the sea-ice thickness variations occur over the whole of Arctic Ocean. The AMV/PMV can speed up or slow down the multidecadal sea-ice loss rates over the Labrador Sea, Greenland Sea, Barents Sea, Bering Sea, and Sea of Okhotsk by more than 50%. The AMV and PMV can affect Arctic sea ice through anomalous atmospheric and oceanic energy convergence, anomalous low cloud cover and the related downward longwave radiation, and changes in sea-ice motion. Anomalous atmospheric energy convergence/divergence contributes to the summertime Arctic sea ice variability. Anomalous horizontal and vertical oceanic heat convergence/divergence is the dominant thermodynamic factor for the wintertime Arctic sea ice variability; it can not only cause sea ice variations but also be an effect of sea ice variations. During DJF, the AMV- and PMV-related low cloud cover anomalies can affect downward longwave radiation and hence influence the Arctic surface air temperature and sea-ice concentration and thickness. The AMV- or PMV-induced sea-ice drift contributes to the DJF Arctic sea ice variations as much as the thermodynamic contributors, but it contributes less (more) to the JJA Arctic sea ice variability than the thermodynamic processes over the marginal Arctic Ocean (the central Arctic Ocean near the Beaufort Sea). The model simulations confirm a positive feedback process: the reduction (increase) of sea ice is further amplified through the enhanced (weakened) air-sea heat fluxes, which warms (cools) near-surface air and the lower-troposphere, resulting in further sea ice loss (increase). Results also suggest that the AMV- and PMV-induced atmospheric circulation anomalies project onto the Arctic Dipole and the Arctic Oscillation and hence have the potential to influence Arctic sea ice indirectly.
Recommended Citation
He, Zhaoxiangrui, "The impacts of Atlantic and Pacific sea surface temperature variability on South American and Arctic climate" (2023). Legacy Theses & Dissertations (2009 - 2024). 3147.
https://scholarsarchive.library.albany.edu/legacy-etd/3147
