Parameterized Convection Model Research Articles

Surface exposure constraints on the mantle water budget

Abstract Mantle water content estimates range from 0.5 to 15 oceans of water. Its evolution is even more unclear. Rapid degassing during mantle solidification likely released much of the water to the surface, initially flooding Earth. However, evidence for subaerial land from at least 3.5 Ga means that much of this water must have been rapidly cycled back into the mantle. Here, we used a parameterized convection model and hypsometric curve to assess how much water could have been taken into the mantle and still satisfy evidence for subaerial land. Even if only the highest peaks were exposed, the initial ocean must have been less than 1.5 current oceans to explain subaerial exposure throughout most of Earth history. Today, this implies any water in the mantle >0.5 oceans must be primordial and has been isolated from the convecting mantle for most of Earth’s history.

Effects on early monsoon rainfall in West Africa due to recent deforestation in a convection-permitting ensemble

Abstract. Tropical deforestation can have a significant effect on climate, but research attention has been directed mostly towards Amazonian deforestation. The southern part of West Africa (a region dependent on rain-fed agriculture and vulnerable to droughts and flooding) has seen significant deforestation since the 1950s. Many previous tropical deforestation studies have used idealized and exaggerated deforestation scenarios and parameterized convection models. In this study we estimate for the first time realistic historical deforestation from the Land-Use Harmonization dataset in West Africa and simulate the impacts in a 5 d ensemble forecast in June using a convection-permitting regional climate model. We find that sensible heat flux increases at the expense of latent heat flux in most deforested regions, and rainfall increases by an average of 8.4 % over deforested pixels from 18:00–06:00 UTC, whereas changes are much less pronounced during the day. Over large areas of deforestation approx. 300 km inland (e.g. west Guinea) the roughness-length and thermally enhanced convergence during the afternoon and evening occurs over the deforested areas resulting in increases in rainfall with little impact from reduced daytime humidity. In areas of coastal deforestation (e.g. Côte d'Ivoire), increased winds drive the sea breeze convection inland, resulting in evening rainfall reductions over the deforested area but increases further inland, in line with observations. We suggest our results would not be replicated in parameterized convection models, which are known to struggle with capturing peak convective activity in the late afternoon and long-lived nocturnal rainfall and with reproducing observed surface–rainfall feedbacks.

Climate Change Impacts on Extreme Rainfall in Eastern Africa in a Convection-Permitting Climate Model

Abstract Climate change is expected to increase the frequency and intensity of rainfall extremes. Understanding future changes in rainfall is necessary for adaptation planning. Eastern Africa is vulnerable to rainfall extremes because of low adaptive capacity and high future population growth. Convection-permitting climate models have been found to better represent moderate (yearly) rainfall extremes than parameterized convection models, but there is limited analysis of rare extremes that occur less frequently than once per year. These events often have the largest socioeconomic impacts. We use extreme value theory and regional frequency analysis to quantify rare rainfall extremes over East Africa in a convection-permitting climate model (CP4A). We compare the results with its parameterized counterpart (P25), the Coordinated Regional Climate Downscaling Experiment for the African region (CORDEX-Africa) ensemble, and observations to understand how the convection parameterization impacts the results. We find that CP4A better matches observations than the parameterized models. With climate change, we find the parameterized convection models have unrealistically high changes in the shape parameter of the extreme value distribution, which controls the tail behavior (i.e., the most extreme events), leading to large increases in return levels of events with a return period of >20 years. This suggests that parameterized convection models may not be suitable for looking at relative changes in rare rainfall events with climate change and that convection-permitting models should be preferred for this type of work. With the more realistic CP4A, RCP8.5 end-of-century climate change leads to 1-in-100-yr events becoming 1-in-23-yr events, which will necessitate serious adaptation efforts to avoid devastating socioeconomic impacts. Significance Statement We use a new, high-resolution climate model to examine how rare extreme rainfall events in East Africa might change in the future with climate change and compare the results with those from standard-resolution climate models. We find that the standard-resolution models have unrealistically large increases in rainfall for events that occur less frequently than every 20 years. The high-resolution model is more realistic and is required to illustrate possible future changes in rare rainfall extremes. Extreme events will become more common with climate change, and in the more realistic model we show that a 1-in-100-yr event may become a 1-in-23-yr event by the end of the century if greenhouse gas emissions are not significantly reduced.

Contrasting contributions of surface hydrological pathways in convection permitting and parameterised climate simulations over Africa and their feedbacks on the atmosphere

The partitioning of rainfall at the land surface into interception, infiltration and surface runoff plays an important role in the water cycle as it controls the time scale at which water returns to the atmosphere. Rainfall intensity is of crucial importance to this partition. High resolution convection permitting models significantly improve simulated sub-daily rainfall intensity distributions, in particular those associated with convective rainfall in the tropics. Here we compare the land surface hydrological response in a pair of 10-year simulations over an African domain performed using the Met Office Unified Model: a typical configuration using parameterised convection operating at 25 km and the second a high resolution convection permitting simulation at 4.5 km with the parametrized convection switched off. Overall pan-African interception in the convection permitting scheme is 70% lower, whilst surface runoff is 43% higher than the parameterized convection model. These changes are driven by less frequent, but more intense rainfall with a 25% increase in rainfall above 20 mm h−1 in the 4.5 km model. The parameterised scheme has a ~ 50% canopy water contribution to evaporative fraction which is negligible in the convection permitting scheme. Conversely, the convection permitting scheme has higher throughfall and infiltration leading to higher soil moisture in the weeks following rain resulting in a 30–50% decrease in the daytime sensible heat flux. We examine how important the sub-grid rainfall parameterisation in the model is for the differences between the two configurations. We show how, switching a convective parameterisation off can substantially impact land surface behaviour.

The Effect of Explicit Convection on Couplings between Rainfall, Humidity, and Ascent over Africa under Climate Change

AbstractThe Hadley circulation and tropical rain belt are dominant features of African climate. Moist convection provides ascent within the rain belt, but must be parameterized in climate models, limiting predictions. Here, we use a pan-African convection-permitting model (CPM), alongside a parameterized convection model (PCM), to analyze how explicit convection affects the rain belt under climate change. Regarding changes in mean climate, both models project an increase in total column water (TCW), a widespread increase in rainfall, and slowdown of subtropical descent. Regional climate changes are similar for annual mean rainfall but regional changes of ascent typically strengthen less or weaken more in the CPM. Over a land-only meridional transect of the rain belt, the CPM mean rainfall increases less than in the PCM (5% vs 14%) but mean vertical velocity at 500 hPa weakens more (17% vs 10%). These changes mask more fundamental changes in underlying distributions. The decrease in 3-hourly rain frequency and shift from lighter to heavier rainfall are more pronounced in the CPM and accompanied by a shift from weak to strong updrafts with the enhancement of heavy rainfall largely due to these dynamic changes. The CPM has stronger coupling between intense rainfall and higher TCW. This yields a greater increase in rainfall contribution from events with greater TCW, with more rainfall for a given large-scale ascent, and so favors slowing of that ascent. These findings highlight connections between the convective-scale and larger-scale flows and emphasize that limitations of parameterized convection have major implications for planning adaptation to climate change.

Model Improvement via Systematic Investigation of Physics Tendencies

Abstract Deficiencies in forecast models commonly stem from inadequate representation of physical processes; yet, improvement to any single physics component within a model may lead to degradations in other physics components or the model as a whole. In this study, a systematic investigation of physics tendencies is demonstrated to help identify and correct compensating sources of model biases. The model improvement process is illustrated by addressing a commonly known issue in warm-season rainfall forecasts from parameterized convection models: the misrepresentation of the diurnal precipitation cycle over land, especially in its timing. Recent advances in closure assumptions in mass-flux cumulus schemes have made remarkable improvements in this respect. Here, we investigate these improvements in the representation of the diurnal precipitation cycle for a spring period over the United States, and how changes to the cumulus scheme impact the model climate and the behavior of other physics schemes. The modified cumulus scheme improves both the timing of the diurnal precipitation cycle and reduces midtropospheric temperature and moisture biases. However, larger temperature and moisture biases are found in the boundary layer as compared to a predecessor scheme, along with an overamplification of the diurnal precipitation cycle, relative to observations. Guided by a tendency analysis, we find that biases in the diurnal amplitude of the precipitation cycle in our simulations, along with temperature and moisture biases in the boundary layer, originate from the land surface model.

Large-Scale Predictors for Extreme Hourly Precipitation Events in Convection-Permitting Climate Simulations

AbstractMidlatitude extreme precipitation events are caused by well-understood meteorological drivers, such as vertical instability and low pressure systems. In principle, dynamical weather and climate models behave in the same way, although perhaps with the sensitivities to the drivers varying between models. Unlike parameterized convection models (PCMs), convection-permitting models (CPMs) are able to realistically capture subdaily extreme precipitation. CPMs are computationally expensive; being able to diagnose the occurrence of subdaily extreme precipitation from large-scale drivers, with sufficient skill, would allow effective targeting of CPM downscaling simulations. Here the regression relationships are quantified between the occurrence of extreme hourly precipitation events and vertical stability and circulation predictors in southern United Kingdom 1.5-km CPM and 12-km PCM present- and future-climate simulations. Overall, the large-scale predictors demonstrate skill in predicting the occurrence of extreme hourly events in both the 1.5- and 12-km simulations. For the present-climate simulations, extreme occurrences in the 12-km model are less sensitive to vertical stability than in the 1.5-km model, consistent with understanding the limitations of cumulus parameterization. In the future-climate simulations, the regression relationship is more similar between the two models, which may be understood from changes to the large-scale circulation patterns and land surface climate. Overall, regression analysis offers a promising avenue for targeting CPM simulations. The authors also outline which events would be missed by adopting such a targeted approach.

Thermal evolution of an early magma ocean in interaction with the atmosphere: conditions for the condensation of a water ocean

The thermal evolution of magma oceans produced by collision with giant impactors late in accretion is xpected to depend on the composition and structure of the atmosphere through the greenhouse effect of CO2 and H2O released from the magma during its crystallization. We developed a 1D parameterized convection model of a magma ocean coupled with a 1D radiative convective model of the atmosphere. We conducted a parametric study and described the influences of some important parameters such as the Sun-planet distance. Our results suggest that a steam atmosphere delays the end of the magma ocean phase by typically 1 Myr. Water vapor condenses to an ocean after 0.1 Myr, 1.5 Myr and 10 Myr for, respectively, Mars, Earth and Venus. This time would be virtually infinite for an Earth-sized planet located at less than 0.66 AU from the Sun. So there are conditions such as no water ocean is formed on Venus. Moreover, for Mars and Earth, water ocean formation time scales are shorter than typical time gaps between major impacts. This implies that successive water oceans may have developed during accretion, making easier the loss of their atmospheres by impact erosion.

THE INFLUENCE OF PRESSURE-DEPENDENT VISCOSITY ON THE THERMAL EVOLUTION OF SUPER-EARTHS

We study the thermal evolution of super-Earths with a one-dimensional (1D) parameterized convection model that has been adopted to account for a strong pressure dependence of the viscosity. A comparison with a 2D spherical convection model shows that the derived parameterization satisfactorily represents the main characteristics of the thermal evolution of massive rocky planets. We find that the pressure dependence of the viscosity strongly influences the thermal evolution of super-Earths—resulting in a highly sluggish convection regime in the lower mantles of those planets. Depending on the effective activation volume and for cooler initial conditions, we observe with growing planetary mass even the formation of a conductive lid above the core-mantle boundary (CMB), a so-called CMB-lid. For initially molten planets our results suggest no CMB-lids but instead a hot lower mantle and core as well as sluggish lower mantle convection. This implies that the initial interior temperatures, especially in the lower mantle, become crucial for the thermal evolution—the thermostat effect suggested to regulate the interior temperatures in terrestrial planets does not work for massive planets if the viscosity is strongly pressure dependent. The sluggish convection and the potential formation of the CMB-lid reduce the convective vigor throughout the mantle, thereby affecting convective stresses, lithospheric thicknesses, and heat fluxes. The pressure dependence of the viscosity may therefore also strongly affect the propensity of plate tectonics, volcanic activity, and the generation of a magnetic field of super-Earths.

Physical state of the deep interior of the CoRoT-7b exoplanet

AbstractThe present study takes the CoRoT-7b exoplanet as an analogue for massive terrestrial planets to investigate conditions, under which intrinsic magnetic fields could be sustained in liquid cores. We examine the effect of depth-dependent transport parameters (e.g., activation volume of mantle rock) on a planet's thermal structure and the related heat flux across the core mantle boundary. For terrestrial planets more massive than the Earth, our calculations suggest that a substantial part of the lowermost mantle is in a sluggish convective regime, primarily due to pressure effects on viscosity. Hence, we find substantially higher core temperatures than previously reported from parameterized convection models. We also discuss the effect of melting point depression in the presence of impurities (e.g., sulfur) in iron-rich cores and compare corresponding melting relations to the calculated thermal structure. Since impurity effects become less important at the elevated pressure and temperature conditions prevalent in the deep interior of CoRoT-7b, iron-rich cores are likely solid, implying that a self-sustained magnetic field would be absent.

Mechanisms for cessation of magmatic resurfacing on Venus

Proposed mechanisms of resurfacing on Venus about 0.3–1 Gyr ago usually involve either some form of tectonic resurfacing in which Venusian lithosphere is recycled mechanically or magmatic resurfacing where essentially immobile lithosphere is covered by lava. The focus of this study is mechanisms of magmatic resurfacing in the stagnant lid regime of mantle convection. Parameterized convection models suggest that cessation of magmatic resurfacing can occur in several ways. (1) The mantle temperature drops sufficiently such that mantle rising adiabatically does not cross the solidus. (2) The molten layer migrates below the solid/melt density inversion at 250–500 km so that no melt can escape. (3) Sublithospheric small‐scale convection stops and conductive thickening of the lid suppresses melting. In each case, inability of magma to penetrate thickened Venusian lithosphere may play a role. The timing of melt cessation and duration of late stage resurfacing depend on various factors such as mantle rheology, degree of mantle depletion, and depth of resurfacing. The models indicate that the waning stages of resurfacing which result in surface age variations can last ∼0.1–1 Gyr. A general trend of recent lithospheric thickening is also predicted. Some cases exhibit a rapid transition from relatively thin to thick lid as suggested by previous authors.

Influence of a variable thermal conductivity on the thermochemical evolution of Mars

The influences of variations in thermal conductivity due to composition, temperature, and pressure in mantle and crust on the thermochemical evolution of Mars have been examined with parameterized convection models. We use a temperature‐ and pressure‐dependent thermal conductivity in the mantle and column‐averaged values between 1.5 and 4 Wm−1K−1 for the crust (with 2 Wm−1K−1 as the representative value for a basaltic crust). The major characteristic of the models is a slower cooling of the upper mantle with a long‐existing global partial melt zone and a stronger crustal production late in the evolution compared to earlier models which assume a constant thermal conductivity of 4 Wm−1K−1. To explain the observed crustal thickness and the early crustal formation on Mars, the formation of an enriched primordial crust is required. Recent volcanism can be explained by lateral temperature variations that are caused by thickness variations of the low conducting crust. These melt zones which occur underneath thicker than average crust may be the source region for the observed longstanding volcanism in Tharsis and Elysium instead of plumes rising from the core‐mantle boundary. The evolution of the elastic lithosphere is characterized by a transition at which the elastic strength is first provided by the crust and later also by the upper mantle. This finding may explain the estimated elastic lithosphere thickness of less than 16 km in the Noachian and more than 100 km in the Hesperian and Amazonian.

Convective instability in Europa's floating ice shell

Models of the tidally heated, floating ice shell proposed for the jovian satellite Europa generally find shell thicknesses less than 30 km. Past parameterized convection models indicated that such shells are stable against convective overturn, which otherwise ostensibly leads to freezing of the ocean underneath. Here I apply the temperature‐dependent viscosity convection scaling developed by Solomatov and coworkers to the Europan ice shell. The temperature‐dependent properties of ice are linearized about 260 K, as any convective interior should be close to this temperature, with the colder ice forming an essentially passive, stagnant lid. Ice shells ≳ 10 km thick are found to be unstable to convection at their base for melting‐point viscosities of 1013 Pa‐s (as linearized by tidal stresses), if the ice deforms by superplastic creep, but such low viscosities require small grain sizes (<1 mm). This requirement may be met if grain sizes observed in terrestrial polar glaciers can be strain‐rate scaled to Europa. Regardless, convection at the base of the ice shell, if initiated, may not freeze the ocean. Because of tidal heating, a stagnant‐lid regime ice shell is much more dissipative than a conductive shell of the same thickness. Such a shell should thin, not thicken, and the potential exists for further thermal instabilities and runaways.

Convective thermal evolution of the upper mantles of Earth and Venus

On Earth the present‐day rate of heat loss is about twice the heat generation rate; on Venus it is about half. Though this rough balance may be due to a feedback mechanism between mantle temperature and heat loss, it is difficult to see how such a mechanism can occur on timescales of 1 Ga or less when the upper mantle of the Earth is thought to be cooling at about 40°C Ga−1. On Venus a decrease in surface heat flux presumably occurred at the end of the catastrophic resurfacing event at ∼500 Ma. Parameterized convection models relate heat flux to Rayleigh number by the exponent β. Such models using a range of viscosities and values of β from 0.2 to 0.3 show that the effect of a sudden decrease in surface heat flux is to cause an independently convecting upper mantle to increase in temperature by 100–500°C over 1 Ga, whereas, if whole mantle convection occurs, the temperature change is less than 60°C. An increase in mantle temperature of 200°C or more will affect mantle viscosity, lithospheric thickness and melt generation rate, all of which may affect the feedback mechanism.

Regional viscosity variations, small‐scale convection and slope of the depth‐age1/2 curve

Cooling plate and cooling half‐space models are commonly employed to describe the depth‐age1/2 curve of seafloor. These conductive models are not able to reproduce the large range of subsidence rates that are observed with reasonable changes in the physical parameters. We examine the influence of small‐scale convection in the mantle on subsidence rates with a one‐dimensional, parameterized‐convection model. The parameterization is based on a recent experimental study of thermal convection in a fluid with temperature‐dependent viscosity. In our model, convection in the asthenosphere begins at the ridge axis, enhancing heat transfer out of the asthenosphere and increasing the subsidence rate above that for simple conductive cooling. Variations in deformation mechanism, temperature and water content of the mantle cause changes in viscosity that affect the vigor of convection and may be responsible for large variations in subsidence rates. Average subsidence rates can range from ∼250–500 m/Ma1/2 when small‐scale convection is considered.

Recent deformation rates on Venus

Constraints on the recent geological evolution of Venus may be provided by quantitative estimates of the rates of the principal resurfacing processes, volcanism and tectonism. This paper focuses on the latter, using impact craters as strain indicators. The total postimpact tectonic strain lies in the range 0.5–6.5%, which defines a recent mean strain rate of 10−18–10−17 s−1 when divided by the mean surface age. Interpretation of the cratering record as one of pure production requires a decline in resurfacing rates at about 500 Ma (catastrophic resurfacing model). If distributed tectonic resurfacing contributed strongly before that time, as suggested by the widespread occurrence of tessera as inliers, the mean global strain rate must have been at least ∼10−155 S−1, which is also typical of terrestrial active margins. Numerical calculations of the response of the lithosphere to inferred mantle convective forces were performed to test the hypothesis that a decrease in surface strain rate by at least two orders of magnitude could be caused by a steady decline in heat flow over the last billion years. Parameterized convection models predict that the mean global thermal gradient decreases by only about 5 K/km over this time; even with the exponential dependence of viscosity upon temperature, the surface strain rate drops by little more than one order of magnitude. Strongly unsteady cooling and very low thermal gradients today are necessary to satisfy the catastrophic model. An alternative, uniformitarian resurfacing hypothesis holds that Venus is resurfaced in quasi‐random “patches” several hundred kilometers in size that occur in response to changing mantle convection patterns. Under such a model, the observed crater strain distribution indicates that about 1% of the planet's surface is tectonically active at any time. However, this model requires a very weak crustal rheology to achieve surface velocities ∼100 mm/yr appropriate to the required “patch” size. Without well‐developed lateral weak zones, Venus is essentially a one‐plate planet, but one in which the lithosphere is able to respond to topography produced by mantle convection through faulting and limited horizontal movement. The net rate of tectonic activity is logarithmically intermediate between Earth and Mars: about 100 times slower than plate tectonics, but up to 100 times faster than planets where tectonic stress arises largely from lithospheric cooling and contraction.

Mantle devolatilization and convection: Implications for the thermal history of the Earth

The depletion of volatiles from the earth's mantle over its history may have had significant effects on thermal convection, due to the effects of volatiles on silicate rheology. Experimental data suggest that the activation energy for solid‐state creep in a volatile‐free environment is as much as 25% greater than in a volatile‐rich environment, and thus devolatilization can lead to a substantial increase in viscosity. In a devolatilizing system, gradually increasing temperatures (relative to those in a system with volatile‐independent rheology) may be required to maintain moderate viscosities and convection. We investigate this postulated effect by incorporating a time‐dependent increase in activation temperature, simulating progressive mantle devolatilization, in a parameterized convection model. Volatiles significantly affect the earth's thermal evolution when the dependence of mantle rheology upon volatile content is sufficiently strong, the total volatile loss is sufficiently large, and devolatilization is relatively slow and continuous. For models with an increase of 20% in activation energy over earth history, the present‐day ratio of heat production to heat loss (Urey ratio) approaches or exceeds 1.0 when volatile loss is relatively continuous. With rapid early volatile loss, the upper bound on the present‐day Urey ratio is about 0.8 to 0.9, comparable to the limit with volatile‐independent rheology. Present uncertainties in these factors permit an earth that may be close to a steady thermal state or perhaps even warming up.

Modes of mantle convection and the removal of heat from the Earth's interior

Earth thermal histories based on two‐layer and whole‐mantle parameterized convection models have been explored to determine the influence of the mode of mantle convection on the content and distribution of mantle radioactive heat sources and on the temperatures and heat transfer properties of the mantle. Two‐layer mantle convection thermally insulates the lower mantle and is less efficient at heat removal than whole mantle convection. Consequently, two‐layer mantle convection requires a lower mantle depleted of radioactive heat sources contrary to the undepleted lower mantles of recent geochemical models. Whole‐mantle convection removes 3 times more primordial heat from the earth's interior and 6 times more heat from the core in 4.5 billion years than does two‐layer convection. The average mantle volumetric heat generation rates for both modes of convection are comparable to that of a potassium‐depleted chondrite. Thus no potassium is necessary in the core to balance the observed present‐day surface heat flux. However, most of the radioactives are concentrated in the upper mantle for two‐layer convection. Model geotherms for whole‐mantle convection compare well with independent estimates of temperature at various depths, including the core‐mantle boundary, while the lower mantle is too hot for two‐layer convection models. Thermal histories of the upper mantle in two‐layer models are similar to those of whole mantle models. Thermal histories for the lower mantle in two‐layer models are similar to those of the core in whole mantle convection models. This is because the lower mantle is thermally insulated by the upper mantle if it convects separately, just as the core is thermally insulated by the whole mantle. The differences between whole and two‐layer mantle convection are mostly due to the thermal insulation of the lower mantle and the smaller efficiency of heat removal of the upper mantle compared to that of the whole mantle, which is much larger in volume. A whole‐mantle convection model reconciling the need for distinct geochemical reservoirs with the physics of mantle convection is presented.