Towards Generative Machine Learning-based Downscaling for Atmosphere-Surface Coupling in the Bern3D EMIC

The isotopic composition of Si in biogenic silica (BSi), such as opal buried in the oceans' sediments, has changed over time. Paleorecords suggest that the isotopic composition, described in terms of δ30Si, was generally much lower during glacial times than today. There is consensus that this variability is attributable to differing environmental conditions at the respective time of BSi production and sedimentation. The detailed links between environmental conditions and the isotopic composition of BSi in the sediments remain, however, poorly constrained. In this study, we explore the effects of a suite of offset boundary conditions during the Last Glacial Maximum (LGM) on the isotopic composition of BSi archived in sediments in an Earth System Model of intermediate complexity (EMIC). Our model results suggest that a change in the isotopic composition of Si supply to the glacial ocean is sufficient to explain the observed overall low(er) glacial δ30Si in BSi. All other processes explored trigger model responses of either wrong sign or magnitude or are inconsistent with a recent estimate of bottom water oxygenation in the Atlantic Sector of the Southern Ocean. Caveats, mainly associated with generic uncertainties in today's pelagic biogeochemical modules, remain.

Abstract. The surface energy and mass balance of ice sheets strongly depends on the amount of solar radiation absorbed at the surface, which is mainly controlled by the albedo of snow and ice. Here, using an Earth system model of intermediate complexity, we explore the role played by surface albedo for the simulation of glacial cycles. We show that the evolution of the Northern Hemisphere ice sheets over the last glacial cycle is very sensitive to the representation of snow albedo in the model. It is well known that the albedo of snow depends strongly on snow grain size and the content of light-absorbing impurities. Excluding either the snow aging effect or the dust darkening effect on snow albedo leads to an excessive ice build-up during glacial times and consequently to a failure in simulating deglaciation. While the effect of snow grain growth on snow albedo is well constrained, the albedo reduction due to the presence of dust in snow is much more uncertain because the light-absorbing properties of dust vary widely as a function of dust mineral composition. We also show that assuming slightly different optical properties of dust leads to very different ice sheet and climate evolutions in the model. Conversely, ice sheet evolution is less sensitive to the choice of ice albedo in the model. We conclude that a proper representation of snow albedo is a fundamental prerequisite for a successful simulation of glacial cycles.

Mid- to Late Holocene climate change: an overview

Due to the large thermal inertia of water, the oceans are Earth’s largest heat reservoir and the primary regulator of global climate on long timescales. The ocean absorbs, stores and distributes heat through the planet, and shifting ocean circulation patterns can disrupt established climate regimes. Comparably, the cryosphere modulates the Earth’s climate because ice sheets reflect radiation from the Sun and serve as a reservoir of freshwater, thereby modulating sea level, which affects the large human populations that inhabit coastal zones. Thus, ocean and cryosphere are two significant agents of Earth’s climate system and tightly connected by feedbacks on millennial and orbital timescales. Understanding the internal mechanisms that control such feedbacks is crucial for refining numerical and conceptual models and aid public policy decisions to deal with the repercussions of anthropogenic climate change. However, scientific knowledge about long-timescale ocean-cryosphere interactions is hampered by the geographically sparse and relative short length of observational records. The reconstruction of climate patterns on longer timescales requires the production and interpretation of proxy records and are the scope of this thesis. The close proximity between ice sheets and areas of deep-water formation makes the North Atlantic one of the most climate sensitive areas on the planet. There, the transition zone between the subtropical and subpolar gyres is susceptible to latitudinal shifts of oceanic fronts and was reached by icebergs laden with debris during periods of ice-sheet destabilization since the beginning of the Pleistocene. Additionally, subsurface density layers that connect both gyres at depth are the pathway through which low-latitude heat may reach high latitudes and either feed moisture or accelerate the demise of ice sheets. To assess the relationship between the ocean heat transport and the cryosphere over millennial and orbital timescales, two sets of proxy records were produced from deep-sea marine sediment cores retrieved from the transition zone between both gyres in the North Atlantic. The first set of records allowed for the reconstruction of subsurface temperatures and salinity during the Last Glacial Maximum and last glacial termination, a period of ice-sheet demise. During this time interval, the climate of the northern hemisphere went through several abrupt shifts accompanied by equally abrupt changes in oceanic circulation, that resulted in a sea-level rise of ca. 120 m above the glacial baseline. A compilation of new and published proxy records showed that subsurface warming preceded the deposition of debris transported by icebergs, implicating that northward heat transport triggered or accelerated the demise of glacial ice sheets. A second set of proxy records produced from materials of the same location was used to reconstruct subsurface temperatures at mid-latitudes and oceanic circulation patterns between mid and high latitudes during the Mid-Pleistocene Transition. This interval was characterized by the transition from 40-kyr to 100-kyr glacial-interglacial cyclicity of the Late Pleistocene and was period of significant ice-sheet enlargement. The produced data suggested a prominent role for low-latitude forcing in driving the northward transport of heat when cold surface temperatures and low atmospheric CO2 would have starved ice sheets of moisture and hampered their growth. Additionally, modern-like gradients between mid- and high-latitude oxygen isotope records from the surface, intermediate and deep ocean, indicate the establishment of the Late Pleistocene ocean circulation during this second interval. In ensemble, the data produced within the framework of this thesis were used to suggest a new mechanism of northward heat transport that is closely coupled to the fate of northern hemisphere ice sheets during the Middle and Late Pleistocene. Furthermore, the new data suggest that the causal relationship between northward heat transport and deep-water formation may need to be revaluated to better constrain climate and oceanic circulation in future high-CO2 model scenarios.

Atmospheric CO2 and polar ice volume have been strongly coupled over the past 805,000 years. However, the prior extent of coupling, during times of lower-amplitude ice-volume variability, is unknown because continuous high-resolution CO2 records are lacking. We reconstructed the past 1,460,000 years of atmospheric CO2 (~1,700 year sample resolution) by taking advantage of the unique relationship between CO2 concentration and leaf-wax δ13C resulting from changes in the extent of C3 and C4 vegetation in East India. Notably, reconstructed interglacial CO2 concentrations were lower before the transition to large volume variability during the mid-Pleistocene Transition (900,000 years ago). Prior to the mid-Pleistocene Transition, CO2 exhibited a secular trend similar to that of deep-ocean carbon isotopes. At orbital time scales, phase analysis indicates that the CO2 lead relative to ice volume changed to a lag during the mid-Pleistocene Transition. Combined, these findings suggest that deep-ocean circulation controlled the long-term CO2 trend, and that interaction between CO2, continental ice and deep-ocean circulation was reorganized during the mid-Pleistocene Transition, involving a decrease in carbon storage in the deep Pacific.

During the Mid-Pleistocene Transition (MPT; 1,200-800 kya), Earth's orbitally paced ice age cycles intensified, lengthened from ∼40,000 (∼40 ky) to ∼100 ky, and became distinctly asymmetrical. Testing hypotheses that implicate changing atmospheric CO2 levels as a driver of the MPT has proven difficult with available observations. Here, we use orbitally resolved, boron isotope CO2 data to show that the glacial to interglacial CO2 difference increased from ∼43 to ∼75 μatm across the MPT, mainly because of lower glacial CO2 levels. Through carbon cycle modeling, we attribute this decline primarily to the initiation of substantive dust-borne iron fertilization of the Southern Ocean during peak glacial stages. We also observe a twofold steepening of the relationship between sea level and CO2-related climate forcing that is suggestive of a change in the dynamics that govern ice sheet stability, such as that expected from the removal of subglacial regolith or interhemispheric ice sheet phase-locking. We argue that neither ice sheet dynamics nor CO2 change in isolation can explain the MPT. Instead, we infer that the MPT was initiated by a change in ice sheet dynamics and that longer and deeper post-MPT ice ages were sustained by carbon cycle feedbacks related to dust fertilization of the Southern Ocean as a consequence of larger ice sheets.

<p>A new simple approach inspired by MacAyeal (1979) to explain the time-asymmetric ‘saw-toothed’ shape and 100,000-year quasi-period of glacial-interglacial cycles since the Middle Pleistocene Transition, is presented. Using a simple model with fast-slow dynamics, the global ice volume is taken to be a function of two independently varying parameters, the solar insolation and ‘alpha’, a secondary control parameter, the study of which is the focus this research. The steady state of the model is a partially folded surface in three-dimensional space where insolation, ‘alpha’, and global ice volume are orthogonal axes. The pleated surface allows for the gradual increase and sudden decrease in ice volume that is observed in the paleoclimate record. To derive a time series of global ice volume, the Euler integration method is used, producing a time series which replicates the ‘saw-toothed’ pattern of glacial cycles in the late Pleistocene. The second control parameter, ‘alpha’, is proposed to be related to internal dynamics of the climate system, such as ice sheet dynamics.</p><p> </p><p>Reference</p><p>D. R.  MacAyeal, ‘A Catastrophe Model of the Paleoclimate Record’ , Journal of Glaciology , Volume 24 , Issue 90 , 1979 , pp. 245 – 257.</p>

Indian-Asian monsoon has oscillated between warm/wet interglacial periods and cool/dry glacial periods with periodicities closely linked to variations in Earth’s orbital parameters. However, processes that control wet versus dry, i.e. aridity cyclical periods on the orbital time-scale in the low latitudes of the Indian-Asian continent remain poorly understood because records over millions of years are scarce. The sedimentary record from International Ocean Discovery Program (IODP) Expedition 359 provides a well-preserved, high-resolution, continuous archive of lithogenic input from the Maldives reflecting on low-latitude aridity cycles. Variability within the lithogenic component of sedimentary deposits of the Maldives results from changes in monsoon-controlled sedimentary sources. Here, we present X-ray fluorescence (XRF) core-scanning results from IODP Site U1467 for the past two million years, allowing full investigation of orbital periodicities. We specifically use the Fe/K as a terrestrial climate proxy reflecting on wet versus dry conditions in the source areas of the Indian-Asian landmass, or from further afield. The Fe/K record shows orbitally forced cycles reflecting on changes in the relative importance of aeolian (stronger winter monsoon) during glacial periods versus fluvial supply (stronger summer monsoon) during interglacial periods. For our chronology, we tuned the Fe/K cycles to precessional insolation changes, linking Fe/K maxima/minima to insolation minima/maxima with zero phase lag. Wavelet and spectral analyses of the Fe/K record show increased dominance of the 100 kyr cycles after the Mid Pleistocene Transition (MPT) at 1.25 Ma in tandem with the global ice volume benthic δ18O data (LR04 record). In contrast to the LR04 record, the Fe/K profile resolves 100-kyr-like cycles around the 130 kyr frequency band in the interval from 1.25 to 2 million years. These 100-kyr-like cycles likely form by bundling of two or three obliquity cycles, indicating that low-latitude Indian-Asian climate variability reflects on increased tilt sensitivity to regional eccentricity insolation changes (pacing tilt cycles) prior to the MPT. The implication of appearance of the 100 kyr cycles in the LR04 and the Fe/K records since the MPT suggests strengthening of a climate link between the low and high latitudes during this period of climate transition.

Vegetation response to obliquity and precession forcing during the Mid-Pleistocene Transition in Western Mediterranean region (ODP site 976)

Widely used MRI methods show brain morphology both in vivo and ex vivo at very high resolution. Many of these methods (e.g., T2*-weighted imaging, phase-sensitive imaging, or susceptibility-weighted imaging) are sensitive to local magnetic susceptibility gradients produced by subtle variations in tissue composition. However, the spectral resolution of commonly used methods is limited to maintain reasonable run-time combined with very high spatial resolution. Here, the authors report on data acquisition at increased spectral resolution, with 3-dimensional high spectral and spatial resolution MRI, in order to analyze subtle variations in water proton resonance frequency and lineshape that reflect local anatomy. The resulting information compliments previous studies based on T2* and resonance frequency. The proton free induction decay was sampled at high resolution and Fourier transformed to produce a high-resolution water spectrum for each image voxel in a 3D volume. Data were acquired using a multigradient echo pulse sequence (i.e., echo-planar spectroscopic imaging) with a spatial resolution of 50 × 50 × 70 μm(3) and spectral resolution of 3.5 Hz. Data were analyzed in the spectral domain, and images were produced from the various Fourier components of the water resonance. This allowed precise measurement of local variations in water resonance frequency and lineshape, at the expense of significantly increased run time (16-24 h). High contrast T2*-weighted images were produced from the peak of the water resonance (peak height image), revealing a high degree of anatomical detail, specifically in the hippocampus and cerebellum. In images produced from Fourier components of the water resonance at -7.0 Hz from the peak, the contrast between deep white matter tracts and the surrounding tissue is the reverse of the contrast in water peak height images. This indicates the presence of a shoulder in the water resonance that is not present at +7.0 Hz and may be specific to white matter anatomy. Moreover, a frequency shift of 6.76 ± 0.55 Hz was measured between the molecular and granular layers of the cerebellum. This shift is demonstrated in corresponding spectra; water peaks from voxels in the molecular and granular layers are consistently 2 bins apart (7.0 Hz, as dictated by the spectral resolution) from one another. High spectral and spatial resolution MR imaging has the potential to accurately measure the changes in the water resonance in small voxels. This information can guide optimization and interpretation of more commonly used, more rapid imaging methods that depend on image contrast produced by local susceptibility gradients. In addition, with improved sampling methods, high spectral and spatial resolution data could be acquired in reasonable run times, and used for in vivo scans to increase sensitivity to variations in local susceptibility.

Irrigated crop production in the Texas high plains (THP) is dependent on water extracted from the Ogallala Aquifer, an area suffering from sever water shortage. Water management in this area is therefore highly important. Thermal satellite imagery at high temporal (∼daily) and high spatial (∼100 m) resolutions could provide important surface boundary conditions for vegetation stress and water use monitoring, mainly through energy balance models such as DisALEXI. At present, however, no satellite platform collects such high spatiotemporal resolution data. The objective of this study is to examine the utility of an image sharpening technique (TsHARP) for retrieving land surface temperature at high spatial resolution (down to 60 m) from moderate spatial resolution (1 km) imagery, which is typically available at higher (∼daily) temporal frequency. A simulated sharpening experiment was applied to Landsat 7 imagery collected over the THP in September 2002 to examine its utility over both agricultural and natural vegetation cover. The Landsat thermal image was aggregated to 960 m resolution and then sharpened to its native resolution of 60 m and to various intermediate resolutions. The algorithm did not provide any measurable improvement in estimating high‐resolution temperature distributions over natural land cover. In contrast, TsHARP was shown to retrieve high‐resolution temperature information with good accuracy over much of the agricultural area within the scene. However, in recently irrigated fields, TsHARP could not reproduce the temperature patterns. Therefore we conclude that TsHARP is not an adequate substitute for 100‐m‐scale observations afforded by the current Landsat platforms. Should the thermal imager be removed from follow‐on Landsat platforms, we will lose valuable capacity to monitor water use at the field scale, particularly in many agricultural regions where the typical field size is ∼100 × 100 m. In this scenario, sharpened thermal imagery from instruments like MODIS or VIIRS would be the suboptimal alternative.

The ∼100 k.y. cyclicity of the late Pleistocene ice ages started during the mid-Pleistocene transition (MPT), as ice sheets became larger and persisted for longer. The climate system feedbacks responsible for introducing this nonlinear ice sheet response to orbital variations in insolation remain uncertain. Here we present benthic foraminiferal stable isotope (δ18O, δ13C) and trace metal records (Cd/Ca, B/Ca, U/Ca) from Deep Sea Drilling Project Site 607 in the North Atlantic. During the onset of the MPT, glacial-interglacial changes in δ13C values are associated with changes in nutrient content and carbonate saturation state, consistent with a change in water mass at our site from a nutrient-poor northern source during interglacial intervals to a nutrient-rich, corrosive southern source during glacial intervals. The respired carbon content of glacial Atlantic deep water increased across the MPT. Increased dominance of corrosive bottom waters during glacial intervals would have raised mean ocean alkalinity and lowered atmospheric pCO2. The amplitude of glacial-interglacial changes in δ13C increased across the MPT, but this was not mirrored by changes in nutrient content. We interpret this in terms of air-sea CO2 exchange effects, which changed the δ13C signature of dissolved inorganic carbon in the deep water mass source regions. Increased sea ice cover or ocean stratification during glacial times may have reduced CO2 outgassing in the Southern Ocean, providing an additional mechanism for reducing glacial atmospheric pCO2. Conversely, following the establishment of the ∼100 k.y. glacial cycles, δ13C of interglacial northern-sourced waters increased, perhaps reflecting reduced invasion of CO2 into the North Atlantic following the MPT.

Hyperspectral images (HSI) have high spectral and low spatial resolutions. However, multispectral images (MSI) usually have low spectral and high spatial resolutions. In various applications HSI with high spectral and spatial resolutions are required. In this paper, a new method for spatial resolution enhancement of HSI using high resolution MSI based on sparse coding and linear spectral unmixing (SCLSU) is introduced. In the proposed method (SCLSU), high spectral resolution features of HSI and high spatial resolution features of MSI are fused. In this case, the sparse representation of some high resolution MSI and linear spectral unmixing (LSU) model of HSI and MSI is simultaneously used in order to construct high resolution HSI (HRHSI). The fusion process of HSI and MSI is formulated as an ill-posed inverse problem. It is solved by the Split Augmented Lagrangian Shrinkage Algorithm (SALSA) and an orthogonal matching pursuit (OMP) algorithm. Finally, the proposed algorithm is applied to the Hyperion and ALI datasets. Compared with the other state-of-the-art algorithms such as Coupled Nonnegative Matrix Factorization (CNMF) and local spectral unmixing, the SCLSU has significantly increased the spatial resolution and in addition the spectral content of HSI is well maintained.

Most modern satellites can provide two types of images: 1) panchromatic (PAN) image and 2) multispectral (MS) image. The former has high spatial resolution and low spectral resolution, while the latter has high spectral resolution and low spatial resolution. To obtain images with both high spectral and spatial resolution, pansharpening has emerged to fuse the spatial information of the PAN image and the spectral information of the MS image. However, most pansharpening methods fail to preserve spatial and spectral information simultaneously. In this article, we propose a framelet-based convolutional neural network (CNN) for pansharpening which makes it possible to pursue both high spectral and high spatial resolution. Our network consists of three subnetworks: 1) feature embedding net; 2) feature fusion net; and 3) framelet prediction net. Different from conventional CNN methods directly inferring high-resolution MS images, our approach learns to predict their framelet coefficients from available PAN and MS images. The introduction of multilevel feature aggregation and hybrid residual connection makes full use of spatial information of PAN image and spectral information of MS image. Quantitative and qualitative experiments at reduced- and full-resolution demonstrate that the proposed method achieves more appealing results than other state-of-the-art pansharpening methods. The source code and trained models are available at https://github.com/TingMAC/FrMLNet.

Momentum‐resolved electron energy loss spectroscopy (MREELS) probes the momentum (q) dependence (dispersion) of energy losses from characteristic excitations such as excitons, plasmon, and interband excitations thus probing bandstructures, indirect excitations, and dipole‐forbidden excitations. 1 However, the ability to obtain momentum‐resolved spectra both at high spatial and momentum resolutions has remained difficult. As such, layer‐resolved momentum spectra from ultra‐thin films, heterostructures as well as from nanostructures has not been widely reported. We present an experimental approach that enables the acquisition of momentum resolved spectra at high spatial resolution down to 2 nm using a nano‐beam electron diffraction approach. 2 Through this approach we have obtained momentum resolved spectra from individual, differently‐oriented nano‐domains in an ultra thin (12 nm) PrNiO 3 layer as well as spectra from different positions in a LaNiO 3 thin film (70 nm). Figure 1(a) displays a Z‐contrast image of the LaNiO 3 thin layer grown on a LaSrAlO 4 substrate. A nano‐beam electron diffraction pattern from the LaNiO 3 layer is shown in Figure 1(b). A slit is used to select the 200, ‐200 set of diffraction spots (shown by the red rectangle) allowing electrons that have been scattered to certain scattering angles into the spectrometer. This corresponds to the Г‐X direction of the Brillouin zone in the pseudo‐cubic symmetry of LaNiO 3 . The resulting image (ω‐q map) shown in figure 1(c) displays energy losses as function of momentum transfer along the Г‐X (q// [100] ) direction in the reciprocal space. The presented approach will enable the acquisition of momentum resolved spectra from nano‐structured materials, thin films, interfaces, surfaces, and heterostructures at high spatial, energy, and momentum resolutions.