Anthropogenic climate change is driving shifts in vegetation communities. These shifts are projected to increase in frequency and extent as climate change intensifies into the future. Trees are long-lived and, in the absence of disturbance, can tolerate years to decades of climate conditions for which they are maladapted. A disturbance, however, can trigger a sudden shift in vegetation type where “legacy” forests are maladapted. Identifying the level of forest maladaptation to climate and likelihood of disturbance could help land managers anticipate sudden vegetation shifts as they plan for the future. As part of the NASA CMS Project, we implemented an Environmental Evaluation Modeling System (EEMS) model to evaluate the risk for sudden forest shift in the western conterminous United States. To determine where forests will be maladapted under future conditions, we simulated vegetation without fire using the MC2 dynamic global vegetation model (DGVM) with 20 different climate scenarios. Vegetation type departure between historical and future periods is used as a proxy for the risk of vegetation shift due to maladaptation. The greater the departure, the greater the risk. To limit our analysis to actual forested areas, we used LandFire forest landcover from the U.S. Departments of Agriculture and Interior. To quantify disturbance risk, the EEMS model uses a disturbance dataset produced as part of the NASA CMS project. In addition to quantifying the risk of sudden vegetation shift, the EEMS model also provides a standalone data layer of forest maladaptation useful for decision support as well as the spatial data layers for each node in the EEMS model. To conclude, we discuss how the model can be combined with other models – e.g. species distribution models and economic models – to further inform land management decisions.

Serpentinization ubiquitously affects ultramafic rocks that interact with water, with strong implications for the origin of ancient microbial ecosystems, the chemical budget of the ocean, and the rheology of the oceanic crust. The increase in volume associated with this reaction, and its consequences for reaction progress, have been debated for over a century. Serpentine minerals are ~40% more voluminous than olivine, which suggests that fully serpentinized peridotite should have negative, or at least very low, porosity. Recent studies have proposed that self-propagating, reaction-driven fracturing can facilitate serpentinization reactions, but the nanoscale mechanisms by which fluid is transported through serpentinized fractures in order to react with fresh olivine surfaces remain poorly understood. To address this issue, we studied a sample of serpentinized harzburgite collected during ODP expedition 209 at site 1274, using Focused Ion Beam – Scanning Electron Microscopy (FIB-SEM)-assisted tomography. We specifically targeted the interface between the serpentinized peridotite and the unaltered primary mineralogy. The resultant images illustrate the presence of nanopores within the serpentine alteration products, primarily at the interface with the unreplaced minerals. Importantly, no pores were observed in the serpentine away from the grain boundaries with olivine, suggesting that these nanopores form during the initial stage of reaction, and then disappear with further reaction progress. We argue that the observed nanoporosity is an intrinsic feature of the serpentinization reaction, and that its transient presence during serpentinization is vital for facilitating reaction progress. We suggest that the transient nature of the pores arises from the opposing kinetics and thermodynamics of the replacement reaction. The former promotes the formation of voids that enhance the advective transport of fluids to the reaction front, while the latter drives the reduction of pore space by means of recrystallization of the serpentine aggregates and minimization of the interfacial energy.

The “Laramide-style” uplifts of North America—characterized by blocks of Proterozoic-Archean basement that were exhumed along reverse faults within the Cordilleran foreland basin—are widely interpreted to be a result of flat-slab subduction of oceanic lithosphere beneath the continent. Despite this general consensus, the causal mechanisms of basement-cored uplifts remain unclear. Assessment of the hotly debated geodynamic models that relate flat-slab subduction and upper crustal deformation hinge on the availability of accurate estimates for the timing of exhumation of Laramide uplifts. A major problem with current models is that they do not incorporate timing constraints from the Laramide region of central Montana. This region represents the northernmost extent of Laramide deformation and timing constraints are critically needed to enhance our understanding of how stress is transmitted inboard during flat-slab subduction. We present the first low-temperature thermochronological ages from the Little Belt Mountains (LBM) of central Montana, which is the northernmost Laramide-style uplift with exposed basement gneisses. Apatite fission track ages ranging from ca. 90-70 Ma suggest that the core of the LBM was exhumed through the 120-60°C apatite partial annealing zone in the Late Cretaceous. These ages corroborate recent low-T thermochronology and sedimentological work that propose an earlier than previously recognized onset of Laramide deformation in southwestern Montana and eastern Idaho (>80 Ma), but additional data are needed to reduce the uncertainty between ages and ascertain the exhumation history of the LBM. To this end, we will integrate new low-temperature thermochronology (<150°C) and associated thermal history models in order to constrain the Cretaceous tectonic evolution of the LBM and the extent of Laramide deformation in the western USA.

The equatorial electrojet (EEJ) is an important manifestation of ionospheric electrodynamics. Day-to-day changes of the EEJ result from E-region dynamo processes that are primarily driven by highly variable atmospheric waves propagating up from the lower and middle atmosphere. Progress has been made in our understanding that upward propagating tides are one of the major contributors to the day-to-day variability in the EEJ, however current models are limited in their ability to capture the vertical propagation of tides from the lower and middle atmosphere to the upper atmosphere due to difficulties to adequately represent many processes that influence it. In this study, we thus propose a new data-driven approach to modeling day-to-day variability by taking advantage of widely available ground-based magnetic field measurements. The new approach based on an ensemble transform adjustment method is applied to the Thermosphere-Ionosphere-Electrodynamics General Circulation Model (TIE-GCM) lower boundary conditions (LBCs) at about 97 km altitude in order to make the model’s tidal characteristics to be more consistent with observed magnetic perturbations associated with the EEJ. In this method, TIE-GCM ensemble simulations are driven by high-latitude ionospheric convection and auroral particle precipitation patterns specified by the AMGeO and by atmospheric waves and tides based on MERRA meteorological reanalysis. As part of forward modeling, the 3D Dynamo electrodynamic module is used to calculate magnetic perturbations on the ground and at low Earth orbit altitudes. A detailed analysis of the 21-day period from March 1 to 22, 2009 has shown that the modeled EEJ with the LBCs adjusted using ground-based magnetic perturbation data improves the agreement of the model to independent magnetic field observations from CHAMP. The use of routinely available ground-based magnetometer data to constrain the TIE-GCM LBCs could provide an opportunity to investigate how day-to-day tidal variability drives equatorial electrodynamics variability.

To address some challenging biological and environmental problems, scientists at the Environmental Molecular Sciences Laboratory (EMSL) at Pacific Northwest National Laboratory (PNNL) have been pairing scientific capabilities at their home institution with the bright, tunable energy x-rays and high-resolution instruments at several national synchrotron user facilities. Synchrotron techniques can help answer a variety questions, and two examples featuring synchrotron-based structural and chemical imaging in bio/geochemical systems will be presented. In the first example, x-ray fluorescence spectromicroscopy and microtomography were used to visualize and better understand phosphorous uptake in poplar trees. Endophyte-promoted phosphorous uptake was seen inside poplar roots, where the fixated phosphorous appeared to be in the form of an organic phosphate. Analysis of the tomography data showing increased root mass for the plants inoculated with the endophytes supported the picture of increased nutrient uptake in those plants. These results along with in-house proteomics characterization point to the biological relevance of the symbiosis between endophytes and the host plant. In the second example, scanning transmission x-ray microscopy (STXM) combined with x-ray absorption near edge structure (XANES) helped us investigate soil mineral – soil organic matter (SOM) interactions in an alkaline soil from Washington state. Ca mineral–organic associations were found to be predominant which may play a critical role in the stabilization/degradation of SOM and mineral. Micro- and nanoscale characterization of the chemical state of both Ca from the mineral and C from the organic matter are crucial for understanding such stabilization mechanisms as well as soil nutrient dynamics.

Volcanic ash detection is important for effective emergency management. Ash emitted by volcanoes is injected at high altitudes into the atmosphere, dispersed and driven by wind and extended for kilometres away from volcanoes. It can affect communities in different ways including health, agriculture, the environment and air traffic. Thus, detecting and tracking the volcanic ash clouds using satellite images can help civil protection authorities and volcanic ash advisory centres respond rapidly to a volcanic event. Satellite based detection of volcanic ash has traditionally entailed the implementation of radiances, reflectances, brightness temperatures (BT), brightness temperature differences (BTD), beta ratios and cloud objects tracking tools. The aim of this study is to test machine learning algorithms on satellite data in order to identify volcanic ash in the atmosphere as an attempt to enhance the space-borne ash detection capabilities especially in ambiguous cases, where image classification is hard, such as when ash, clear land and cirrus clouds coexist within a scene. We used a set of 10 satellite images acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) on board the satellites Terra and Aqua just after the 4 June 2011 eruption of the Puyehue Cordon Caulle Volcanic Complex to train and test two supervised classification algorithms: K-Nearest Neighbours and Support Vector Machine, for volcanic ash detection in the atmosphere. We first built a database that consists of 217,859 pixels identified and labelled, on the basis of visual interpretation of color composites, according to seven classes: volcanic ash above the emission centre, land and sea, meteorological clouds above land and sea, and clear sky above land and sea. For each of the seven classes, sample pixel values were retrieved for six features, which are given by the BT at 11𝝁m and the BTDs between 11 and 12 𝝁m, 11 and 3.9 𝝁m, 11 and 6.7 𝝁m, 11 and 8.5 𝝁m and a BTD that combines the BTs measured at 11, 12 and 8.5 𝝁m. We show preliminary results on statistical analysis of the training datasets and performance of both algorithms for volcanic ash detection above land and sea including verification tests with MODIS datasets that cover the April 2015 eruption of Volcan Calbuco (Chile).

In this study, we revisit the role of curvature in modifying frontal stability. We first consider the statement “fq < 0 implies potential for instability”, where f is the Coriolis parameter and q is the Ertel potential vorticity (PV). This is true for any inviscid baroclinic flow. It is also evident in the transition of a governing equation for circulation within a front from elliptic to hyperbolic form as the discriminant changes sign. However, for curved fronts, an additional scale factor enters the discriminant owing to conservation of absolute angular momentum, L, leading to Solberg’s (1936) generalization of the Rayleigh criterion. In non-dimensional form, this expression also generalizes the classical instability criterion of Hoskins (1974) by accounting for centrifugal forces: modification of the front’s vertical shear and stratification owing to curvature tilts the absolute vorticity vector away from its thermal wind state and, in an effort to conserve the product of non-dimensional PV (q’) and absolute angular momentum (L’), this alters Rossby and Richardson numbers permitted for stable flow. The criterion, Φ’=L’q’ < 0, is then investigated in non-dimensional parameter space representative of low-Richardson-number vortices. An interesting outcome is that, for Richardson numbers near one, anticyclonic flows increase in q’, while cyclonic flows decrease in q’. Though stabilization is muted for anticyclones (owing to multiplication by L’), the de-stabilization of cyclones is robust, and may help to explain an observed asymmetry in the distribution of submesoscale coherent vortices in the global ocean.

Post-seismic deformation following large earthquakes offers insights into the rheology of the lithosphere and upper asthenosphere. The Mojave, in southern California, is one of the best studied regions on Earth, yet key questions, about fault slip rates and rheological heterogeneity, remain unanswered. Unprecedented geodetic coverage of the 2019 Ridgecrest earthquakes provides an opportunity to test whether rheological models developed for the Mojave, from the Landers, Hector Mine and El Major Cucapah earthquakes, are applicable north of the Garlock fault, and to place bounds on the effects of local rheological heterogeneties associated with the Coso volcanic field. This volcanic field, which is located to the NW of the Mw7.1 rupture trace, is a region of high heat flow and geothermal activity. The locally high temperatures in the Coso volcanic field are likely to be associated with low viscosities compared to the surrounding regions, and high pore pressures due to the hydrothermal activity. The aftershock sequence associated with the Ridgecrest earthquakes shows a notable absence of large magnitude earthquakes in this region. We use variational Bayesian independent component analysis to isolate postseismic deformation in GPS time series around the earthquakes. We present models of the possible poroelastic, afterslip and viscoelastic response driven by coseismic stress changes in the July 2019 Ridgecrest earthquakes and investigate the possible effect of the Coso volcanic field. By modelling a series of different afterslip geometries, and viscoelastic rheologies we identify features of the GPS- and InSAR-derived surface deformation which are diagnostic of different post-seismic mechanisms and rheological heterogeneities.

To better resolve the bed elevation of Antarctica, we present DeepBedMap - a deep learning method that produces realistic Antarctic bed topography from multiple remote sensing data inputs. Our super-resolution deep convolutional neural network model is trained on scattered regions in Antarctica where high resolution (250 m) groundtruth bed elevation grids are available, and then used to generate high resolution bed topography in less well surveyed areas. DeepBedMap takes in a low resolution (1000 m) BEDMAP2 dataset alongside other high spatial resolution inputs such as ice surface elevation, velocity and snow accumulation to generate a four times upsampled (250 m) bed topography map even in the absence of ice-thickness data from direct seismic or ice-penetrating radar surveys. Our DeepBedMap model is based on an Enhanced Super Resolution Generative Adversarial Network architecture that is adapted to minimize per-pixel elevation errors while producing realistic topography. We show that DeepBedMap offers a more realistic topographic roughness profile compared to a standard bicubic interpolated BEDMAP2, and also run model inversions to compare the basal traction of our DeepBedMap_DEM with other bed elevation models.

Recently, statistical machine learning and deep learning methods have been widely explored for corn yield prediction. Though successful, machine learning models generated within a specific spatial domain often lose their validity when directly applied to new regions. To address this issue, we designed an unsupervised adaptive domain adversarial neural network (ADANN). Specifically, through domain adversarial training, the ADANN model reduced the impact of domain shift by projecting data from different domains into the same subspace. Also, the ADANN model was designed to be trained in an adaptive way, which guaranteed the model can learn the domain-invariant features and perform accurate yield prediction simultaneously. Informative variables including time-series vegetation indices and sequential weather observations were first collected from multiple data sources and aggregated to the county level. Then, we trained the ADANN model with the extracted features and corresponding reported county-level corn yield from the U.S. Department of Agriculture (USDA). Finally, the trained model was evaluated in four testing years 2016-2019. The U.S. corn belt was used as the study area and counties under study were grouped into two diverse ecological regions. The experimental results showed that the developed ADANN model had better performance than three other state-of-the-art machine learning models in both local experiments (train and test in the same region) and transfer experiments (train and test in different regions). As the first study using adversarial learning for crop yield prediction, this research demonstrates a novel solution for improving model transferability on crop yield prediction.

Test-particle and Fokker-Planck simulations of the energetic electrons trapped in the terrestrial magnetic field are used to study the outer radiation belt electron flux losses during and after geomagnetic storms. We compare the results with ESA PROBA-V observations of electron flux at LEO and with those from the NASA Van Allen Probes mostly at MEO. We find that loss mechanisms of trapped electrons can be very different depending on the geomagnetic activity. Dropouts are visible at all energy during each storm from both satellites. Test-particle simulations show that the Dst (Disturbance storm time) effect during the main phase of a geomagnetic storm results in an outward radial drift and a deceleration of the electrons. This outward drift motion is energy independent, pitch angle dependent, and represents a significant distance of about 1 L-shell at L=5 for moderate storms. At fixed L-shell, this causes a sharp decay of the LEO precipitating flux. The Dst effect, associated with magnetopause shadowing and radial diffusion can explain the main characteristics of outer radiation belt electron dropouts appearing at the beginning of storms. These instantaneous dropouts have to be distinguished from the gradual scattering that depopulates the slot region and the outer belt after storms. Fokker-Planck simulations with event-driven diffusion coefficients at high temporal resolution reproduce the slot formation and the gradual loss in the outer belt. The typical energy-dependence of these losses leads to the absence of scattering for relativistic and ultra-relativistic electrons in the outer belt, oppositely to dropouts. This work has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 870437 for the SafeSpace (Radiation Belt Environmental Indicators for the Safety of Space Assets) project, and for the PITHIA-NRF project (funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101007599).

The U.S. and much of the world sit on the cusp of an electrification revolution – a moment driven largely by the need to reduce the emission of greenhouse gases to limit the impacts of anthropogenic climate change. The electrify everything movement aims to transition combustion-powered sectors into technologies powered solely by electricity, with the idea that the electric grid – currently comprised of a mix of combustion, renewable, and nuclear generation units – will become cleaner and greener over time. Studies indicate that electrifying high-efficiency devices, appliances, and vehicles will reduce greenhouse gas emissions regardless of the grid’s composition, however ancillary air quality benefits and tradeoffs remain poorly resolved, particularly at impact- and equity-relevant scales. Here, I use a fine-scale CONUS-wide climate and air quality co-benefit and tradeoff analysis framework (i.e., 4 km2 SMOKE-CMAQ-WRF simulations) to assess sustainable climate solutions. Analyses utilize emission scenarios that account for increased grid demand and uncertainties in grid evolution, simulate the interaction of meteorological and chemical processes, characterize changes in greenhouse gases and air pollutants, and assess economic, social, and public health consequences of sustainable transitions over two key, yet methodologically disparate, residential/commercial sectors: transportation: via the replacement of internal combustion vehicles with electric vehicles and lighting: via the replacement of low efficiency bulbs with high-efficiency LEDs.

Droughts are occurring with increased frequency and duration in tropical rainforests due to climate change, having a significant impact on soil C dynamics. The role of microbes as drivers of changing C flow, particularly in relation to volatile organic compound (VOC) cycling, remains largely unknown. Here, we aimed to characterize microbial responses to drought using an integrative, multiple ‘omics approach, and hypothesized that microbial communities will adapt by altering their C allocation strategies. Specifically, during pre-drought, primary metabolic pathways will be more active with microbes using C towards growth, whereas during drought, microbes will divert C to secondary metabolite (including VOC) production in response to stress. To test this, we conducted an ecosystem-wide 66-day drought experiment in the tropical rainforest biome at Biosphere 2, a glass- and steel-enclosed facility near Tucson, AZ. To track carbon allocation by microbes, we injected C1 or C2 position-specific 13C-pyruvate solution into a 25 cm2 region within a soil flux chamber collar (n=6 locations) and measured C isotope ratios of VOC and CO2 emissions. Soil was collected at 0, 6, and 48 hours after pyruvate addition to examine responses in soil metatranscriptomics, metagenomics, and metabolomics (1H nuclear magnetic resonance [NMR] and Fourier-transform ion cyclotron resonance [FTICR]). Our results indicated that 13CO2 (primarily emitted from C1-13C-pyruvate) fluxes decreased during drought, indicating diminished microbial activity. 13C-VOCs (primarily emitted from C2-13C-pyruvate) fluxes also differed between pre-drought and drought. Furthermore, drought-induced increases in activity of VOC-producing metabolic pathways, including acetate and acetone biosynthesis, were evident, as inferred from volatilome, metabolome, and metatranscriptome data. Overall, these results indicate that integration of multiple ‘omics datasets reveal specific impacts of drought on microbial activity affecting carbon flow in the tropical rainforest soil.

Understanding the linkages between between climatic and surface properties that influence water uptake and loss by vegetation is essential for understanding the impact of drought on dryland regions. The Normalized Difference Vegetation Index (NDVI) is a common metric used to identify vegetation condition across LULC types. Here we employ empirical dynamic modeling (EDM) to forecast NDVI changes for savannas, grasslands, and croplands across East Africa at a dekadal (10-day) time scale using satellite-derived environmental forcing variables. The model relies on state space reconstruction with lagged coordinate embedding of multiple time series observations to recover the dynamic environmental system that links vegetation dynamics to environmental forcing. We apply convergent cross mapping based on Takens’ Theorem to detect the impact of landcover on directional causal interactions and time delays between driving (e.g. LST, rainfall) and response variables (NDVI). The model brings to light that certain regions are highly consistent in their trajectories and therefore easier to project while other regions are more dispersive and thus more difficult to determine anomalies. In terms of land cover, we are able to make projections with high accuracy for grasslands out to 6 months ahead while croplands and savannas show reduced forecast skill overall and prove less useful after 3 months. The use of historical seasonal NDVI patterns to diagnose the manner by which landcover and land use determine climate-land surface couplings provides a means for defining critical areas of inquiry related to the impacts of future change, particularly the expansion of agricultural areas. In addition, the EDM approach provides a robust means for creating short term vegetation forecasts across LULC types in East Africa. These predictions can assist relief organizations in advising drought management, declaring food security classifications and providing early response to famine.

We propose a ‘system-of-systems’—an integrated web of SpWx stations and state-of-the-art modeling facilities to enable a transformative advance in Space Weather nowcasting and forecasting. The Space Weather Aggregated Network of Systems (SWANS) will enable space situational awareness for end-users invested in spaceflight operations, infrastructure risk mitigation, and future human endeavors in space exploration while profoundly transforming Heliophysics research by 2050 or earlier.

On the 20th April 2020, the end of the first strict lockdown in Morocco, 2403410 cases of the corona Virus were confirmed globally. The number of Morocco confirmed cases attended 2990 and 143 deaths were recorded. Due to the pandemic, all avoidable activities in the country were prohibited since the kingdom announced the general lockdown on 20th March 2020. This study aims at comparing the air quality status in Casablanca and Marrakech from Morocco, before the pandemic and during the confinement to show whether COVID-19 compelled lockdown may have saved lives by restraining air pollution than by preventing infection. We used the difference-in-difference and the Theil and Sen non-parametric approaches for univariate time series. We defined the before quarantine period as between the 16th February and the 19th March and the during quarantine as between the 20th March and 20th April. We assessed changes in air quality during vs. before the quarantine period in 2020 and compared these with corresponding changes in the same lunar calendar periods in 2016-2019. Then we calculated the avoided cause-specific mortality attributable to the decreases in NO2 and PM2.5 based on the concentration-response functions from previous studies. We found that NO2 dropped by -12 μg/m³ in Casablanca and -7 μg/m³ in Marrakech. PM2·5 dropped by -18 μg/m³ in Casablanca and -14 μg/m³ in Marrakech. CO dropped by -0.04 mg/m³ in Casablanca and -0.12 mg/m³ in Marrakech. This air pollution reduction had created human health benefits and had reduced mortality and saved lives mainly from cardiovascular diseases. Our results are in complete agreement with the worldwide studies. Yet, they should be interpreted carefully because of the potential common impacts NO2 and PM2·5 may have on health. Further investigation may be undertaken to explore the reduction in the concentrations of industry-related pollutants.

We present seven new transit light curves of the hot-Jupiter TrES-5b observed with the 2-m Himalayan Chandra Telescope, Hanle, India during 2016-2020 to look for the transit timing variation (TTV). For this study, total 59 transit light curves have been considered which include seven light curves from our new observations, 37 from the literature, and 15 from the Exoplanet Transit Database (ETD). In order to have homogeneously derived mid-transit times for precise TTV study, all the light curves were analyzed with a uniform procedure using the Transit Analysis Package (TAP). By fitting a linear ephemeris model to the mid-transit time data, we derive new transit ephemeris with $\chi^{2}_{red}= 2.43$ for TrES-5b, which is consistent and even more precise than the previous results. The best-fit ephemeris with $\chi^{2}_{red} >1$ indicates that the linear ephemeris model does not represent the transit time data well. To look for the periodicity in the timing residuals of linear ephemeris model, a generalized Lomb-Scargle periodogram is computed. However, we have not found any significant periodicity in the timing residuals. The highest peak power obtained in the periodogram at the frequency of 0.005232 rad/period has the false alarm probability (FAP) of 12%, which is found below from the threshold values (i.e. FAP =5% and 1%). This result enable us to conclude that the additional planet might not be present in the TrES-5 system. To confirm this, the further high-precision follow-up observation of the TrES-5 system would be required.

Saturn’s moon Enceladus harbours a global subsurface ocean beneath its icy crust. Tidal dissipation within the moon’s core generates a substantial amount of heat which leads to ocean convection. Understanding this transport of mass and heat within the ocean is key to understand exchange processes between core, ocean and ice shell. Previous studies of ocean convection assume oceans of constant thickness and constant superadiabatic temperature gradients between the inner core and outer ice shell. However, observations indicate ocean thickness variations of up to ~20 km from equator to pole and heterogenous tidal dissipation within the core likely results in latitude-dependent temperature gradients. In this study, we analyse how heterogeneities affect circulation patterns and heat transfer in an Enceladan ocean using three-dimensional direct numerical simulations. We focus on a degree-2 meridional thickness profile with varying degree of thermal forcing to analyse the different rotating convection regimes and compare them with those of an ocean of constant thickness. Our results show that the non-uniform ocean gives rise to a latitudinally variable Rayleigh number, causing a potential decoupling of rotating convection regimes experienced locally within the ocean. As in the constant ocean thickness scenario, different regimes of convection exist, which depend on the relative influence of rotation. With increasing thermal forcing, convection moves from being restricted to equatorial regions to filling the whole fluid domain. Heat transport efficiency, as measured by the local Nusselt number, is different for the uniform and non-uniform cases and depends on the convection regime. Nusselt-Rayleigh relations are similar to those obtained for an ocean of uniform thickness, but the relevant Rayleigh number depends on the region of the domain that is convecting.