Artificial neural network assisted Bayesian calibration of climate models

We demonstrate and validate a Bayesian approach to model calibration applicable to computationally expensive General Circulation Models (GCMs) that includes a posterior estimate of the intrinsic structural error of the model. Bayesian artificial neural networks (BANNs) are trained with output from a GCM and used as emulators of the full model to allow a computationally efficient Markov Chain Monte Carlo (MCMC) sampling of the posterior for the GCM parameters calibrated against seasonal climatologies of temperature, pressure, and humidity. We validate the methodology by calibrating to targets produced by a model run with added noise. We then demonstrate a calibration of five GCM parameters against an observational data set. The approach accounts for both parametric and structural uncertainties of the model as well as uncertainties associated with the observational calibration data. This enables the generation of statistically rigorous probabilistic forecasts for future climate states. All calibration experiments are performed with emulators trained using a maximum of one hundred model runs, in accord with typical resource restrictions imposed by computationally expensive models. We conclude by summarizing remaining issues to address in order to create a complete and validated operational methodology for objective calibration of computationally expensive models.     

Detection of waves converted from P to SV in the mantle

A method is described for the detection of P to SV converted waves in the long-period P coda. The procedure involves axis rotation, transformation of records to a standard form and stacking of processed records from events of various epicentral distances. When applied to NORSAR records, the procedure detected converted phases corresponding to the boundaries in the 410–440- and 640–690-km depth ranges.     

Evaluation of the effective cross-section of scattering in the lithosphere

The theory of scattering affirms that, when a wave enters a scattering volume, secondary waves develop having propagation directions which form various angles with the ray of the primary wave. The energy of the scattered wave depends on the effective cross-section of scattering from a unit volume in a unit solid angle. The seismic phase P′·P′ is produced mainly by the scattering of P′₁ in the uppermost layer of the Earth. If the epicentral distances of the observation point of P′·P′ vary between 30 and 70°, the corresponding angles of scattering of P′ vary between 22° and a few degrees. This relationship enables us to evaluate the effective cross-section of scattering as a function of the angle of scattering. The latest world-average estimates of this quantity, which are based on a large number of seismic data, are presented. The values obtained are in good agreement with those calculated for the Gaussian auto-correlation function of the fluctuating part of the refractive index if the correlation interval is close to 13 km. The lower limit of the fractional r.m.s. of velocity variation in the lithosphere is close to 1.6%.     

Deep structure of the East European platform according to seismic data

The paper presents a review and analysis of new seismic data related to the structure of the mantle beneath the East European platform. Analysis of observations of long-range profiles revealed pronounced differences in the structure of the lower lithosphere beneath the Russian plate and the North Caspian coastal depression. The highest P-velocities found at depths around 100 km are in the range 8.4–8.5 km s^?1. Deep structure of the Baltic shield is different from the structures of both these regions. No evidence of azimuthal anisotropy in the upper mantle was found. A distribution of P-velocity in the upper mantle and in the transition zone consistent with accurate travel-time data was determined. The model involves several zones of small and large positive velocity gradients in the upper mantle, rapid increases of velocity near 400 and 640 km depths and an almost constant positive velocity gradient between the 400 and 640 km discontinuities. The depth of the 640 km discontinuity was determined from observations of waves converted from P to SV in the mantle.     

A Bulk Theory For Air Mass Motion Along A High Mountain Ridge

A steady-state, spatial, large-scale, non-linear problemof the air massmotion along an undulating mountain ridge is consideredin the framework of bulktheory. The ridge is assumed to be so high that the airmass cannot top it, and, insteadof the actual ridge itself, a high vertical wall withsinuousities identical to those of theridge is considered. It is assumed that the air massis bounded above by an inversioninterface (idealized inversion layer) overlain by ageostrophic, polytropic, atmospherethat is thermally homogeneous along the horizontaland stably stratified with aconstant geostrophic wind blowing along the meandirection of the ridge. Theinversion strength (potential temperature deficit) isnot constant and considered as anadditional dependent variable. Because of the Earth'srotation effects, the air massflow to the left of the ridge and that to the rightof the ridge differ considerably intheir features. The fact that the characteristictransverse linear scale of the problem(the generalized Rossby radius of deformation) issmall compared with thelongitudinal scale permits making simplificationsthat result in a semi-geostrophicmodel of the boundary-layer type. Then the problemcan be reduced to an ordinarydifferential equation, which admits a closed-formsolution. Analysis of the solutionenables one to deduce some general features ofthe process under investigation suchas, for example, orographic front formation,a transition from sub-critical to super-criticalwind and others.     

Terrane analysis and accretion in North-East Asia

Abstract A terrane map of North-East Asia at 1:5 000 000 scale has been compiled. The map shows terranes of different types and ages accreted to the North-Asian craton in the Mesozoic–Cenozoic, sub-and superterranes, together with post-amalgamation and post-accretion assemblages. The great Kolyma-Omolon superterrane adjoins the north-east craton margin. It is composed of large angular terranes of continental affinity: craton fragments and fragments of the passive continental margin of Siberia, and island arc, oceanic and turbidite terranes that are unconformably overlain by shallow marine Middle-Upper Jurassic deposits. The superterrane resulted from a long subduction of the Paleo-Pacific oceanic crust beneath the Alazeya arc. Its south-west boundary is defined by the Late Jurassic Uyandina-Yasachnaya marginal volcanic arc which was brought about by subduction of the oceanic crust that separated the superterrane from Siberia. According to paleomagnetic evidence the width of the basin is estimated to be 1500–2000 km. Accretion of the superterrane to Siberia is dated to the late Late Jurassic-Neocomian. The north-east superterrane boundary is defined by the Lyakhov-South Anyui suture which extends across southern Chukotka up to Alaska. Collision of the superterrane with the Chukotka shelf terrane is dated to the middle of the Cretaceous. The Okhotsk-Chukotka belt, composed of Albian-Late Cretaceous undeformed continental volcan-ites, defines the Cretaceous margin of North Asia. Terranes eastward of the belt are mainly of oceanic affinity: island arc upon oceanic crust, accretion wedge and turbidite terranes, as well as cratonic terranes and fragments of magmatic arcs on the continental crust and metamorphic terranes of unclear origin and age. The time of their accretion is constrained by post-accretionary volcanic belts that extend parallel to the Okhotsk-Chukotka belt but are displaced to the east: the Maastrichtian-Miocene Kamchatka-Koryak belt and the Eocene-Quaternary Central Kamchatka belt which mark active margins of the continent of corresponding ages.     

Kinematic evolution of the Tethys belt from the Atlantic ocean to the pamirs since the Triassic

We present an updated series of kinematic reconstructions of the major plates around the Tethys from the Atlantic Ocean to the Pamirs between the Early Jurassic and the Present. This set is used elsewhere as a basis for paleogeographic maps of the entire region. The problems related to the positions of the continents in the Lower Triassic are also discussed. No direct analyses of magnetic anomalies and fracture zones in the Atlantic have been made. Rather, all available poles and rotations have been tested in order to eliminate or minimize possible kinematic errors. The reconstructions are shown for nine key geological periods which correspond to well recognized magnetic anomalies, except for the Jurassic-Cretaceous boundary and the Cretaceous-Tertiary boundary which correspond to interpolated positions. Paleolatitudes have been drawn using the study of Westphal et al. (1986). An attempt has been made to take into account the displacements caused by formation of the continental margins and basins by stretching. The resulting relative vector of motions along the northern boundary of the Tethys shows a significant change 80 m.y. ago. Left-lateral motion with compression dominates before whereas right-lateral motion with compression dominates after. To the east, rates of motion vary by a factor of three with time and four maxima can be clearly related to tectonic events in the Late Jurassic, Late Cretaceous, Eocene and post Middle Miocene. To the west, north of Apulia, on the contrary, the motion rate has not changed significantly since the Early Cretaceous and is close to 1 cm/yr as an average. These rather complex adjustments in rates and directions of relative motion are produced in great part through a complex migration of the Africa-Eurasia pole of rotation and seem to be mostly governed by the tectonics of the Tethys plate boundary.     

Stability of iron oxides and their role in the formation of rock magnetism

Thermodynamic conditions (first of all, temperature) are the main dynamic factors in the transformation process of ferrous to ferric iron (TFFI). TFFI usually takes place within a temperature range of 473–843 K (most active at temperatures above 673 K) and does not require presence of the oxidizing agents above 673 K. Analysis of the chemical composition of different rocks and minerals indicates that only for some sedimentary rocks is the relative content of ferrous iron oxide less than its value in magnetite, and this value is minimal for oceanic sediments. The relative content of ferrous iron oxide in oceanic magmatic rocks exceeds this value in continental magmatic rocks and depends on the rate of rock cooling. An investigation of the role of the titanium oxide content of different rocks on stability of ferrous iron oxide against its transformation to ferric iron oxide shows that a significant correlation (r = 0.79) does exist between the relative content of ferrous iron oxide and ratio of TiO₂/Fe₂O₃. Temperature within the solar nebula at location of the Earth was within the temperature range of the TFFI. During the Earth accretion and its early evolution, ferric iron oxide was unstable and most likely did not exist. The first magnetic minerals containing ferric iron could have appeared only after the Earth’s surface had cooled below ∼843 K. The formation of the first Algoma-type banded iron formations could be used as a marker of the Earth’s surface cooling below ∼843 K.     

Coupling coordination analysis with data-driven technology for disaster–economy–ecology system: an empirical study in China

Natural Hazards - Nowadays, frequent meteorological disasters that cause huge economic losses and ecological damages have swept the world. Thus, research investigates how to overcome the adverse...     

A search for evidence of small-scale inhomogeneities in dense cores from line profile analysis

In order to search for intensity fluctuations on the HCN(1–0) and HCO~+(1–0) line profiles, which could arise due to possible small-scale inhomogeneous structure, long-term observations of high-mass star-forming cores S140 and S199 were carried out. The data were processed by the Fourier filtering method. Line temperature fluctuations that exceed the noise level were detected. Assuming the cores consist of a large number of randomly moving small thermal fragments, the total number of fragments is ~ 4 × 10~6 for the region with linear size ~ 0.1 pc in S140 and ~ 10~6 for the region with linear size ~ 0.3 pc in S199. Physical parameters of fragments in S140 were obtained from detailed modeling of the HCN emission in the framework of the clumpy cloud model.