Search

The last interglacial contains information about how climate behaves during a warm period comparable to the current interval. However, there is no consensus on the key signatures and force of the last interglacial climate changes at present. Here we present a high-resolution stalagmite oxygen isotope record from central China, which spans from 123.3 to 116.8 ka. The stalagmite δ18O time-series bears a resemblance to changes in Northern Hemisphere summer insolation before the demise of the last interglacial, like the condition during the Holocene. Superimposed on the long-term δ18O record is a pronounced shift to more positive values at around 121.6 ka, indicating a switch of the Asian monsoon mean state into a weaker condition. This monsoon shift correlates with a decrease in the mean state of the South American Summer Monsoon, and a well-known abrupt mid-Eemian cooling event at high latitudes. These correlations indicate broad-scale features of the climate shift over the warm last interglacial. Another positive δ18O shift of up to 3.5‰ occurred at around 120 ka, representing the demise of the last interglacial Asian monsoon. By comparing the stalagmite δ18O records against changes in ice volume and summer insolation during the last interglacial and recent Holocene, respectively, we conclude that the decreasing Northern Hemisphere summer insolation dominated the rapid demise of the last interglacial in the Asian monsoon region.

In this paper, we propose a point cloud based 3D object detection framework that accounts for both contextual and local information by leveraging multi-receptive field pillars, named as MuRF-Net. Recently, common pipelines can be divided into a voxel-based feature encoder and an object detector. During the feature encoding steps, contextual information is neglected, which is critical for the 3D object detection task. Thus, the encoded features are not suitable to input to the subsequent object detector. To address this challenge, we propose the MuRF-Net with a multi-receptive field voxelization mechanism to capture both contextual and local information. After the voxelization, the voxelized points (pillars) are processed by a feature encoder, and a channel-wise feature reconfiguration module is proposed to combine the features with different receptive fields using a lateral enhanced fusion network. In addition, to handle the increase of memory and computational cost brought by multi-receptive field voxelization, a dynamic voxel encoder is applied taking advantage of the sparseness of the point cloud. Experiments on the KITTI benchmark for both 3D object and Bird's Eye View (BEV) detection tasks on car class are conducted and MuRF-Net achieved the state-of-the-art results compared with other voxel-based methods. Besides, the MuRF-Net can achieve nearly real-time speed with 20Hz.

The trajectory prediction is significant for the decision-making of autonomous driving vehicles. In this paper, we propose a model to predict the trajectories of target agents around an autonomous vehicle. The main idea of our method is considering the history trajectories of the target agent and the influence of surrounding agents on the target agent. To this end, we encode the target agent history trajectories as an attention mask and construct a social map to encode the interactive relationship between the target agent and its surrounding agents. Given a trajectory sequence, the LSTM networks are firstly utilized to extract the features for all agents, based on which the attention mask and social map are formed. Then, the attention mask and social map are fused to get the fusion feature map, which is processed by the social convolution to obtain a fusion feature representation. Finally, this fusion feature is taken as the input of a variable-length LSTM to predict the trajectory of the target agent. We note that the variable-length LSTM enables our model to handle the case that the number of agents in the sensing scope is highly dynamic in traffic scenes. To verify the effectiveness of our method, we widely compare with several methods on a public dataset, achieving a 20% error decrease. In addition, the model satisfies the real-time requirement with the 32 fps.

The empirical mode decomposition (EMD) method is used to re-analyse the high-resolution and precisely-dated stalagmite record from Chinese caves over the past 640 ka. Results show that (1) the variation in the Asian Monsoon can be completely decomposed into ten quasiperiod oscillations, among which the precession and semiprecession band oscillations are the most prominent periodicities, with contribution rates of 31.1% and 30.7%, respectively; (2) the cross-spectrum analysis of the semiprecession component and bi-hemisphere insolation (BHI) are strongly correlated, indicating an amplified response of precipitation and temperature variability to the interhemispheric insolation in the low-latitude regions, thus further affecting the intensity of the Asian Monsoon; (3) on millennial timescales, obvious oscillations at the 5 ka and 1–2 ka bands roughly correspond to the classical Bond and Dansgaard-Oeschger (DO) cycles. Additionally, a strong correlation is found between the detrended stalagmite δ18O records and Ca/Sr sequence from the North Atlantic (especially at the 5 ka band). This result means that the 5 ka cycle is characteristic of the glacial-interglacial cycle since the middle and late Pleistocene and may imply that climate change on the millennial timescale is the result of an interaction between global ice volume and insolation.

The whole story An accurate, precise record of the carbon-14 ( 14 C) content of the atmosphere is important for developing chronologies in climate change, archaeology, and many other disciplines. Cheng et al. provide a record that covers the full range of the 14 C dating method (∼54,000 years), using paired measurements of 14 C/ 12 C and thorium-230 ( 230 Th) ages from two stalagmites from Hulu Cave, China. The advantage of matching absolute 230 Th ages and 14 C/ 12 C allowed the authors to fashion a seamless record from a single source with low uncertainties, particularly in the older sections. Science , this issue p. 1293

Due to the significance of electron and heat transfer in designing the nanoscale semiconductor devices, the transient analysis of electrothermal properties has attracted extensive attention. In this paper, the density-gradient and dual-phase-lag (DPL) equations are first combined to predict the electron and heat transport in nanoscale transistors. The DPL equation is solved with the consideration of the temperature jump boundary condition that dealing with phonon–wall collisions. We have shown that the temporal and spatial distributions of related physical variables can be obtained by self-consistently solving these equations. Furthermore, the spectral element time-domain method is used to discretize these equations. Numerical results of electrothermal properties for both 2-D and 3-D field-effect transistors have been demonstrated to show the robustness and universality of the proposed model. Therefore, the model we proposed can be used with the temporal and spatial distributions, which could be helpful for evaluating the electrothermal performance and computational designing of nanoscale transistors.

Despite the links of Asian monsoon with climates at high northern and southern latitudes, it remains unclear that at which time and to what extent the Asian monsoon variation is dominated by one of the two drivers throughout a Greenland Stadial (GS) to Greenland Interstadial (GI) cycle. Here we provide a Chinese stalagmite δ18O record to study their teleconnections throughout the GS-6 to GI-5.2 cycle. The resemblance between the stalagmite and Greenland records, in timing, duration and abruptness of GI-5.2, supports that the occurrence and termination of GIs are paced by the northern driving force. During the intervals of GI-5.2 and GS-6, however, the Asian monsoon fluctuated concomitantly with variation in temperature over Antarctica, instead of over Greenland. This covariation indicates dominant influences of the Antarctic climate during the climatically stable intervals of stadials and interstadials. This study updates our knowledge on mechanical dynamics of the Asian monsoon change and global climate change throughout a GS to GI cycle.

Autonomous driving is conducted in complex scenarios, which requires to detect 3D objects in real time scenarios as well as accurately track these 3D objects in order to get such information as location, size, trajectory, velocity. MOT (Multi-Object Tracking) performance is heavily dependent on object detection. Once object detection gives false alarms or missing alarms, the multi-object tracking would be automatically influenced. In this paper, we propose a coupling system which combines 3D object detection and multi-object tracking into one framework. We use the tracked objects as a reference in 3D object detection, in order to locate objects, reduce false or missing alarms in a single frame, and weaken the impact of false and missing alarms on the tracking quality. Our method is evaluated on kitti dataset and is proved effective.

In this Letter, a d'Alembert-Schrodinger hybrid method is proposed to analyze the transient interaction between the incident electromagnetic control pulse and the electron. This hybrid method is based on the d'Alembert equation, which describes the propagation of the electromagnetic field and the time-dependent Schrodinger equation, which describes the action of the electron. Moreover, the finite-difference time-domain method is used to solve those equations. In our simulation, using the presented hybrid equations and the control equation of the quantum state, a scheme is presented to design laser pulses to control discrete quantum states in a three-dimensional artificial atom model. Excitingly, the laser pulses have been successfully designed for the perfect four quantum states' transition for the first time. With that, the spatiotemporal distribution for the probability density of an electron wave packet is showed in detail to describe the laser-induced transition process of quantum states.

In the field of autonomous driving, multi-beam light detection and ranging (3D LIDAR) system and global navigation satellite system/integrated inertial navigation system (GNSS/INS) are widely used in high-definition map construction, localization and obstacle detection. As 3D LIDAR system and INS have their own coordinate systems, the calibration of the two mentioned systems is required. In this paper, a novel algorithm for calibrating the coordinate system of 3D LIDAR and INS is proposed, which consists of three parts. The first procedure is to project two point clouds to the world coordinate system based on the initial transform matrix between 3D LIDAR and INS with the real-time data from INS. Then optimal point-to-point correspondences can be found between two frames of point cloud data through registration method. Finally, the loss function is constructed with the sum of the Euclidean distances of the corresponding points and optimized by using perturbation model of Lie algebras, so as to obtain the optimal transform matrix. With different given initial calibration parameters, test results of both simulation and real experiments validate the proposed algorithm and quantify its accuracy and robustness.

This letter presents a field–circuit coupling method for analysis of microwave circuit included with nonlinear physical-model-based semiconductor devices. The electromagnetic field is analyzed by the time-domain volume–surface integral equation (TD-VSIE) and the physical-model-based semiconductor is solved by spectral-element time-domain method. The discrete Newton–Raphson method is employed to solve the proposed field–circuit coupling equations. Compared to the traditional field–circuit coupling method in TD-VSIE, the proposed method can analyze not only the equivalent-model-based semiconductor circuit, but also the physical-model-based one. Finally, numerical results are given to demonstrate the accuracy and efficiency of the proposed method.

The 60-year quasi-periodic variability of the Asian monsoon in the mid-late Holocene has been found to be widespread in instrumental and various terrestrial records. However, whether this period exists in the older interglacial periods remains to be further studied. This paper is based on an 1840-mm-long stalagmite with clear annual laminae (with a growth rate of about 1 mm/yr) from Yongxing Cave, China, that allows for the reconstruction of the Asian monsoon evolution sequence during the period ~123.20–121.21 kyr BP. The stalagmite record closely tracks the 65°N solar insolation changes as a whole, indicating the dominant control of solar insolation on the Asian summer monsoon. After removing the solar insolation signal from the YX108 record, the δ18O residuals (Δδ18O) show 28 multidecadal monsoon cycles. The average amplitude of these fluctuations is larger than 0.8‰, with a duration typically ranging from 50 to 80 yr (about 64 yr on average), which is very similar to the mid-late Holocene Asian monsoon variability. Further, wavelet analysis displays a significant quasi-periodicity of 60 years throughout the studied period, providing further evidence for assessing whether the 60-year quasi-periodic variability of the Asian monsoon found in the Holocene exists in older periods. Our study thus suggests that this quasi-periodic variability not only occurred during the modern period but also existed in the Last Interglacial, and their influence may even be an intrinsic characteristic of the Asian monsoon system during glacial-interglacial cycles. We manifested that the interactions between solar activity and the Atlantic Multidecadal Oscillation may be the main reason for the 60-year quasi-period of the Asian monsoon.

Stable isotope data of precipitation (δ18Op and deuterium excess), drip water (δ18Od), and modern calcite precipitates (δ18Oc and δ13Cc) from Yongxing Cave, central China, are presented, with monthly sampling intervals from June 2013 to September 2016. Moderate correlations between the monthly variation of δ18Op values (from − 11.5 to − 0.7‰) and precipitation amount (r = − 0.59, n = 34, p

Precession is the major orbital control of monsoon intensity, but the monsoonal response to the transition between precessional cycles has yet to be explored. Here, we present an Asian Summer Monsoon history that spans 327.7–280.5 ka BP based on a 230Th-dated and centennially resolved multi-proxy speleothem record from Yongxing Cave, central China. The δ18O profile, a proxy of monsoon intensity, follows the summer insolation changes, with three abrupt shifts occurring at precessional transitions. These shifts are supported by the stalagmite δ13C, grey-level and petrography analysis, indicating a coherent link between the karst processes and monsoon changes. Evidence from the coeval rapid changes in atmospheric CH4 and forest evolution records within dating errors suggests a wider regional signal of monsoon changes. The weak monsoon intervals are temporally consistent with the ice-rafted debris layers in the North Atlantic, indicating a strong coupling of the high- and low-latitude climate systems. We speculate that the abrupt monsoonal shifts are initiated by the south–north shifts of the Intertropical Convergence Zone associated with the Atlantic Meridional Overturning Circulation and further amplified by land cover. In addition, the rates and magnitudes of insolation changes modify monsoon shifts, thus indicating the abrupt monsoonal shift as a result of an interaction of orbital and millennial forcing.

Part 2: Deep Learning; International audience; Lane boundary detection is a key technology for self-driving cars. In this paper, we propose a spatiotemporal, deep learning based lane boundary detection method that can accurately detect lane boundaries under complex weather conditions and traffic scenarios in real time. Our algorithm consists of three parts: (i) inverse perspective transform and lane boundary position estimation using the spatial and temporal constraints of lane boundaries, (ii) convolutional neural networks (CNN) based boundary type classification and position regression, (iii) optimization and lane fitting. Our algorithm is designed to accurately detect lane boundaries and classify line types under a variety of environment conditions in real time. We tested our proposed algorithm on three open- source datasets and also compared the results with other state-of-the-art methods. Experimental results showed that our algorithm achieved high accuracy and robustness for detecting lane boundaries in a variety of scenarios in real time. Besides, we also realized the application of our algorithm on embedded platforms and verified the algorithm’s real-time performance on real self-driving cars.

The regional climate correlation within the Northern Hemisphere in the cold/dry mid-Younger Dryas event (YD) remains elusive. A key to unraveling this issue is sufficient knowledge of the detailed climate variability at the low latitudes. Here we present a high-resolution (3-yr) δ18O record of an annually laminated stalagmite from central China that reveals a detailed Asian monsoon (AM) history from 13.36 to 10.99 ka. The YD in this record is expressed as three phases, characterized by gradual onsets but rapid ends. During the mid-YD, the AM variability exhibited an increasing trend superimposed by three centennial oscillations, well-correlated to changes in Greenland temperatures. These warming/wetting fluctuations show a periodicity of ~ 200 yr, generally in agreement with centennial changes in cosmogenic nuclides indicated by the10Be flux from the Greenland ice. This relationship implies that centennial-scale climate changes during the mid-YD are probably caused by solar output and rapidly transported over broad regions through atmosphere reorganization.

There is increasing evidence that millennial-scale climate variability played an active role on orbital-scale climate changes, but the mechanism for this remains unclear. A 230Th-dated stalagmite δ18O record between 88 and 22 thousand years (ka) ago from Yongxing Cave in central China characterizes changes in Asian monsoon (AM) strength. After removing the 65°N insolation signal from our record, the δ18O residue is strongly anti-phased with Antarctic temperature variability on sub-orbital timescales during the Marine Isotope Stage (MIS) 3. Furthermore, once the ice volume signal from Antarctic ice core records were removed and extrapolated back to the last two glacial-interglacial cycles, we observe a linear relationship for both short- and long-duration events between Asian and Antarctic climate changes. This provides the robust evidence of a link between northern and southern hemisphere climates that operates through changes in atmospheric circulation. We find that the weakest monsoon closely associated with the warmest Antarctic event always occurred during the Terminations. This finding, along with similar shifts in the opal flux record, suggests that millennial-scale events play a key role in driving the deglaciation through positive feedbacks associated with enhanced upwelling and increasing CO2.

Oxygen isotope records from Chinese caves characterize changes in both the Asian monsoon and global climate. Here, using our new speleothem data, we extend the Chinese record to cover the full uranium/thorium dating range, that is, the past 640,000 years. The record’s length and temporal precision allow us to test the idea that insolation changes caused by the Earth’s precession drove the terminations of each of the last seven ice ages as well as the millennia-long intervals of reduced monsoon rainfall associated with each of the terminations. On the basis of our record’s timing, the terminations are separated by four or five precession cycles, supporting the idea that the ‘100,000-year’ ice age cycle is an average of discrete numbers of precession cycles. Furthermore, the suborbital component of monsoon rainfall variability exhibits power in both the precession and obliquity bands, and is nearly in anti-phase with summer boreal insolation. These observations indicate that insolation, in part, sets the pace of the occurrence of millennial-scale events, including those associated with terminations and ‘unfinished terminations’. Records of the Asian monsoon have been extended to 640,000 years ago, and confirm both that the 100,000-year ice age cycle results from integral numbers of precessional cycles and that insolation influences the pacing of major millennial-scale climate events. Prior records of the Asian monsoon have revealed cyclic variations over hundreds of thousands of years, probably driven by variations in insolation caused by the precession of Earth's orbit. Hai Cheng and colleagues now provide a speleothem record from Chinese cave samples that extends earlier records to 640,000 years ago, close to the maximum age possible with uranium/thorium dating. This spectacular record confirms that the characteristic '100,000-year' ice age cycle corresponds to an integral number (four or five) of precession cycles, and that insolation influences millennial-scale variations in monsoon strength.

Nature 534, 640–646 (2016); doi: 10.1038/nature18591 In this Article, the basis for the spectral analysis in Fig. 4 was not clear. The analysis was based upon the past 400,000 years of record, not the full 640,000 years of record. Details in the Methods are correct, but elsewhere the basis is stipulated as the full 640,000 years.