共查询到20条相似文献，搜索用时 890 毫秒
1.
The effect of vegetation on the Younger Dryas (YD) climate is studied by comparing the results of four experiments performed with the ECHAM-4 atmospheric general circulation model (AGCM): (1) modern control climate, (2) simulation with YD boundary conditions, but with modern vegetation, (3 and 4) identical to (2), but with paleo-vegetation. Prescribing paleo-vegetation instead of modern vegetation resulted in temperature anomalies (both positive and negative) of up to 4°C in the Northern Hemisphere mid-latitudes, mainly as an effect of changes in forest cover (change in albedo). Moreover, changes in precipitation and evaporation were found, most notably during December–January–February (DJF) in the tropics and were caused by the replacement of forests by grasslands. These results are consistent with other model studies on the role of vegetation changes on climate and they suggest that it is important in paleoclimate simulation studies to prescribe realistic vegetation types, belonging to the period of interest. However, in our case the addition of YD vegetation did not improve the agreement with proxy data in Europe, as the temperatures were increasing during winter compared to the YD simulation with modern vegetation. It must be noted that this increase was not statistically significant. The model-data mismatch suggests that other factors probably played an important role, such as permafrost and atmospheric dust. We infer that during the last glacial-interglacial transition, the time lag between the first temperature increase and the northward migration of trees, estimated at 500–1000 years, could have delayed the warming of the Eurasian continent. The relatively open vegetation that existed during the early stages of the last glacial-interglacial transition had a relatively high albedo, thus tempering warming up of the Eurasian land surfaces. 相似文献
2.
恒星光谱分类是天文学中一个重要的研究问题.对于已经采集到的海量高维恒星光谱数据的分类,采用模式匹配方法对光谱型分类较为成功,但其缺点在于标准恒星模版之间的差异性在匹配实际观测数据中不能体现出来,尤其是当需要进行光谱型和光度型的二元分类时模版匹配法往往会失败.而采用谱线特征测量的光度型分类强烈地依赖谱线拟合的准确性.为了解决二元分类的问题,介绍了一种基于卷积神经网络的恒星光谱型和光度型分类模型(Classification model of Stellar Spectral type and Luminosity type based on Convolution Neural Network, CSSL CNN).这一模型使用卷积神经网络来提取光谱的特征,通过注意力模块学习到了重要的光谱特征,借助池化操作降低了光谱的维度并压缩了模型参数的数量,使用全连接层来学习特征并对恒星光谱进行分类.实验中使用了大天区面积多目标光纤光谱天文望远镜(Large Sky Area Multi-Object Fiber Spectroscopy Telescope, LAMOST)公开数据集Data Release 5 (DR5,用了其中71282条恒星光谱数据,每条光谱包含了3000多维的特征)对该模型的性能进行验证与评估.实验结果表明,基于卷积神经网络的模型在恒星的光谱型分类上准确率达到92.04%,而基于深度神经网络的模型(Celestial bodies Spectral Classification Model, CSC Model)只有87.54%的准确率; CSSL CNN在恒星的光谱型和光度型二元分类上准确率达到83.91%,而模式匹配方法MKCLASS仅有38.38%的准确率且效率较低. 相似文献
3.
星系的光谱包含其内部恒星的年龄和金属丰度等信息, 从观测光谱数据中测量这些信息对于深入了解星系的形成和演化至关重要. LAMOST (Large Sky Area Multi-Object Fiber Spectroscopic Telescope)巡天发布了大量的星系光谱, 这些高维光谱与它们的物理参数之间存在着高度的非线性关系. 而深度学习适合于处理多维、海量的非线性数据, 因此基于深度学习技术构建了一个8个卷积层$+$4个池化层$+$1个全连接层的卷积神经网络, 对LAMOST Data Release 7 (DR7)星系的年龄和金属丰度进行自动估计. 实验结果表明, 使用卷积神经网络通过星系光谱预测的星族参数与传统方法基本一致, 误差在0.18dex以内, 并且随着光谱信噪比的增大, 预测误差越来越小. 实验还对比了卷积神经网络与随机森林回归模型、深度神经网络的参数测量结果, 结果表明卷积神经网络的结果优于其他两种回归模型. 相似文献
4.
针对BP (Back Propagation)神经网络模型预测卫星钟差中权值和阈值的最优化问题, 提出了基于遗传算法优化的BP神经网络卫星钟差短期预报模型, 给出了遗传算法优化BP神经网络的基本思想、具体方法和实施步骤. 为验证该优化模型的有效性和可行性, 利用北斗卫星导航系统(BeiDou navigation satellite system, BDS)卫星钟差数据进行钟差预报精度分析, 并将其与灰色模型(GM(1,1))和BP神经网络模型预报的结果比较分析. 结果表明: 该模型在短期钟差预报中具有较好的精度, 优于GM(1,1)模型和BP神经网络模型. 相似文献
5.
Soma Mukherjee Ujjwal Bhattacharya S. K. Parui Ranjan Gupta Ravi K. Gulati 《Astrophysics and Space Science》1996,239(2):361-373
In this paper we present an application of an artificial neural network model based on a multi-layered backpropagation algorithm for spectral classification of UV data from the International Ultraviolet Explorer (IUE) low dispersion spectra reference atlas. The model used is similar to that of von Hippel et al. (1994), and is found to reduce the classification error as compared to the recently reported results on the same data set (Gulati et al. 1994b). The improved version of the network is much simpler in structure and the training time is reduced by a factor of almost 20. Such networks will prove very useful in efficient classification of large databases
Subject headings: neural networks, stellar spectra, classification 相似文献
6.
The purpose of this study is to perform a high-resolution general circulation model (GCM) experiment to quantify the sensitivity of regional climate to change in vegetation around the Mediterranean basin, corresponding to vegetation change during the Roman Classical Period (RCP), about 2000 years BP. First, an RCP vegetation distribution based on fossil pollen maps and historical records was defined. Second, the RCP vegetation inferred from palynology and other proxies was converted to the 12 vegetation types required by the biosphere model implemented in the GCM. The albedo change due to the change in vegetation significantly alters the atmospheric circulation over northern Africa and the Mediterranean. The consequences of this change involve a northward shift of the ITCZ in the African continent and a coupled circulation between northwestern Africa and the Mediterranean Sea. A large increase of precipitation occurs over the Sahel, the Nile valley and northwestern Africa. A smaller increase of precipitation occurs also over the Iberian Peninsula and the region corresponding to the south of the Caucasus range (Armenia). The increase of precipitation over northern Africa, the Iberian Peninsula and the Armenian region are consistent with the pollen, historical and geographical data. These results suggest that deforestation around the Mediterranean during the last 2000 years contributed to the dryness of the current climate. 相似文献
7.
8.
Stellar Spectral Classification using Principal Component Analysis and Artificial Neural Networks 总被引：1，自引：0，他引：1
Harinder P.Singh Ravi K.Gulati RanjanGupta 《Monthly notices of the Royal Astronomical Society》1998,295(2):312-318
A fast and robust method of classifying a library of optical stellar spectra for O to M type stars is presented. The method employs, as tools: (1) principal component analysis (PCA) for reducing the dimensionality of the data and (2) multilayer back propagation network (MBPN) based artificial neural network (ANN) scheme to automate the process of classification. We are able to reduce the dimensionality of the original spectral data to very few components by using PCA and are able to successfully reconstruct the original spectra. A number of NN architectures are used to classify the library of test spectra. Performance of ANN with this reduced dimension shows that the library can be classified to accuracies similar to those achieved by Gulati et al. but with less computational load. Furthermore, the data compression is so efficient that the NN scheme successfully classifies to the desired accuracy for a wide range of architectures. The procedure will greatly improve our capabilities in handling and analysing large spectral data bases of the future. 相似文献
9.
《Physics and Chemistry of the Earth, Part C: Solar, Terrestrial & Planetary Science》2000,25(4):281-285
Radio waves of a wide range of frequencies from very low frequency (VLF) to high frequency (HF), (broadly 3 to 30 MHz) can be propagated to great distances via the ionosphere.Since the largest variability occurs in the F-region the objective of this paper is to demonstrate a neural network model with the backpropagation algorithm which is designed to forecast the foF2 values of the highly nonlinear ionosphere up to 24 hours in advance. In other words, the model forecasts all values from 1 to 24 hours ahead.By using foF2 data for three European Ionospheric stations this neural network based model can forecast foF2 values both in time and in space for those three stations. The model seems promising for practical work since the root mean square errors involved are within reasonable limits. 相似文献
10.
Traditional methods for predicting the change in length of day (LOD change) are mainly based on some linear models, such as the least square model and autoregression model, etc. However, the LOD change comprises complicated non-linear factors and the prediction effect of the linear models is always not so ideal. Thus, a kind of non-linear neural network — general regression neural network (GRNN) model is tried to make the prediction of the LOD change and the result is compared with the predicted results obtained by taking advantage of the BP (back propagation) neural network model and other models. The comparison result shows that the application of the GRNN to the prediction of the LOD change is highly effective and feasible. 相似文献
11.
12.
《Physics and Chemistry of the Earth, Part C: Solar, Terrestrial & Planetary Science》1999,24(4):385-388
Three different models of electron concentration height profile have been compared to profiles obtained from ionograms at the Warsaw station. The electron density profiles are obtained from ionograms by inversion methods. These models use as an input the standard ionospheric characteristics. Some of the models are intended to compute monthly median profiles. To obtain the instantaneous profile the values of needed parameters are taken from the simultaneously constructed instantaneous maps of ionospheric parameters. For instantaneous maps construction, two techniques are used: the kriging technique with modifications concerning ionospheric behaviour, where deviations of measurements from monthly median maps are used, and a fitting method where median maps are updated with measurements. The comparison is done for the COST 238 PRIME height profile model, a local model based on a modified Rush model, and a model where an artificial neural network technique is aaplied to time series of profiles. The usefulness of the average representative height profile constructed from a set of instantaneous profiles is discussed. 相似文献
13.
射电望远镜天线伺服控制系统中的非线性特性, 对系统动力学特性辨识有着显著的影响, 会提高辨识难度, 增加辨识模型的复杂程度. 系统非线性特性的测量与补偿也会增加系统辨识工作量. 针对上述问题, 提出了一种基于非线性采样数据的线性重构方法, 用于动力学特性建模. 通过提取原采样数据的相位与幅值, 对受到噪声与非线性畸变影响的系统采样数据进行线性重构, 降低待辨识模型的复杂度. 搭建了半实物实验平台, 以平台实际采样为基础, 重构线性数据, 利用奇异值法与自回归神经网络评估并辨识平台动力学模型. 实验结果表明, 建模数据奇异值拐点从100阶下降至40阶, 仅用10个神经网络节点200次训练即实现了模型辨识. 相似文献
14.
One of the main obstacles for extracting the Cosmic Microwave Background (CMB) signal from observations in the mm-submm range is the foreground contamination by emission from Galactic components: mainly synchrotron, free-free and thermal dust emission. Due to the statistical nature of the intrinsic CMB signal it is essential to minimize the systematic errors in the CMB temperature determinations. Following the available knowledge of the spectral behavior of the Galactic foregrounds simple power law-like spectra have been assumed. The feasibility of using a simple neural network for extracting the CMB temperature signal from the combined signal CMB and the foregrounds has been investigated. As a specific example, we have analysed simulated data, as expected from the ESA Planck CMB mission. A simple multilayer perceptron neural network with 2 hidden layers can provide temperature estimates over more than 80 per cent of the sky that are to a high degree uncorrelated with the foreground signals. A single network will be able to cover the dynamic range of the Planck noise level over the entire sky. 相似文献
15.
针对目前从海量的快速射电暴(Fast Radio Burst, FRB)候选体中人工筛选FRB事件难以为继的现状,提出了一种基于卷积神经网络(Convolutional Neural Networks, CNN)的FRB候选体分类方法.首先,通过真实的观测数据和仿真FRB组成训练和测试样本集.其次,建立了二输入的深度卷积神经网络模型,并对其进行训练、测试和优化,获取FRB候选体分类器.最后,利用来自脉冲星的单脉冲数据对该分类器的有效性和性能进行了验证.实验结果表明,该方法可以快速从大量候选体中准确识别出单脉冲事件,极大地提高了FRB候选体的处理速率和效率. 相似文献
16.
The global climate–vegetation model HadSM3_TRIFFID has been used to estimate the equilibrium states of climate and vegetation with pre-industrial and last glacial boundary conditions. The present study focuses on the evaluation of the terrestrial biosphere component (TRIFFID) and its response to changes in climate and CO2 concentration. We also show how, by means of a diagnosis of the distribution of plant functional types according to climate parameters (soil temperature, winter temperature, growing-degree days, precipitation), it is possible to get better insights into the strengths and weaknesses of the biosphere model by reference to field knowledge of ecosystems.The model exhibits profound changes between the vegetation distribution at the Last Glacial Maximum and today that are generally consistent with palaeoclimate data, such as the disappearance of the Siberian boreal forest (taiga), an increase in shrub cover in Europe and an increase of the subtropical desert area. The effective equatorial and sub-tropical tree area is reduced by 18%. There is also an increase in cover of wooded species in North-Western Africa and in Mexico. The analysis of bioclimatic relationships turns out to be an efficient method to infer the contributions of different climatic factors to vegetation changes, both at high latitudes, where the position of the boreal treeline appears in this model to be more directly constrained by the water stress than by summer temperature, and in semi-humid areas where the contributions of temperature and precipitation changes may partly compensate each other. Our study also confirms the major contribution of the decrease in CO2 to environmental changes and carbon storage through its selective impact on gross primary productivity of C3 and C4 plants and a reduction by 25% of water-use efficiency. Specifically, the reduction in CO2 concentration increases the amount of precipitation necessary to sustain at least 20% of grass fraction by 50 mm/year; the corresponding threshold for trees is increased by about 150 mm/year. As a consequence, a reduction in CO2 concentration considerably widens the climatic range where grasses and shrubs dominate. 相似文献
17.
The sensitivity of ECHAM4/ML to a double CO2 scenario for the Late Miocene and the comparison to terrestrial proxy data 总被引：2，自引：2，他引：0
Anke Steppuhn Arne Micheels Angela A. Bruch Dieter Uhl Torsten Utescher Volker Mosbrugger 《Global and Planetary Change》2007,57(3-4):189-212
For the Tortonian, Steppuhn et al. [Steppuhn, A., Micheels, A., Geiger, G., Mosbrugger, V., 2006. Reconstructing the Late Miocene climate and oceanic heat flux using the AGCM ECHAM4 coupled to a mixed-layer ocean model with adjusted flux correction. Palaeogeography, Palaeoclimatology, Palaeoecology, 238, 399–423] perform a model simulation which considers a generally lower palaeorography, a weaker ocean heat transport and an atmospheric CO2 concentration of 353 ppm. The Tortonian simulation of Steppuhn et al. [Steppuhn, A., Micheels, A., Geiger, G., Mosbrugger, V., 2006. Reconstructing the Late Miocene climate and oceanic heat flux using the AGCM ECHAM4 coupled to a mixed-layer ocean model with adjusted flux correction. Palaeogeography, Palaeoclimatology, Palaeoecology, 238, 399–423] demonstrates some realistic trends: the high latitudes are warmer than today and the meridional temperature gradient is reduced. However, the Tortonian run also indicates some insufficiencies such as too cool mid-latitudes which can be due to an underestimated pCO2 in the atmosphere. As a sensitivity study, we perform a further model experiment for which we additionally increase the atmospheric carbon dioxide (700 ppm). According to this CO2 sensitivity experiment, we find a global warming and a globally more intense water cycle as compared to the previous Tortonian run. Particularly the high latitudes are warmer in the Tortonian CO2 sensitivity run which leads to a lower amount of Arctic sea ice and a reduced equator-to-pole temperature difference. Our Tortonian CO2 sensitivity study basically agrees with results from recent climate model experiments which consider an increase of CO2 during the next century (e.g. [Cubasch, U., Meehl, G.A., Boer, G.J., Stouffer, R.J., Dix, M., Noda, A., Senior, C.A., Raper, S., Yap, K.S., 2001. Projections of Future Climate Change. In: Houghton, J.T., Y. Ding, D.J. Griggs, M. Noguer, P.J. van der Linden, X. Dai, K. Maskell, C.A. Johnson (eds.), Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, 525–582]) suggesting that the climatic response on a higher atmospheric CO2 concentration is almost independent from the different settings of boundary conditions (Tortonian versus today). To validate the Tortonian model simulations, we perform a quantitative comparison with terrestrial proxy data. This comparison demonstrates that the Tortonian CO2 sensitivity experiment tends to be more realistic than the previous Tortonian simulation by Steppuhn et al. [Steppuhn, A., Micheels, A., Geiger, G., Mosbrugger, V., 2006. Reconstructing the Late Miocene climate and oceanic heat flux using the AGCM ECHAM4 coupled to a mixed-layer ocean model with adjusted flux correction. Palaeogeography, Palaeoclimatology, Palaeoecology, 238, 399–423]. However, a high carbon dioxide concentration of 700 ppm is questionable for the Late Miocene, and it cannot explain shortcomings of our Tortonian run with ‘normal’ CO2. In order to fully understand the Late Miocene climate, further model experiments should also consider the palaeovegetation. 相似文献
18.
Application of artificial neural networks to the prediction of dust storms in Northwest China 总被引：2，自引：0，他引：2
Artificial neural networks (ANN) are non-linear mapping structures analogous to the functioning of the human brain. In this study, we take the ANN approach to model and predict the occurrence of dust storms in Northwest China, by using a combination of daily mean meteorological measurements and dust storm occurrence. The performance of the ANN model in simulating dust storm occurrences is compared with a stepwise regression model. The correlation coefficients between the observed and the estimated dust storm occurrences obtained from the neural network procedure are found to be significantly higher than those obtained from the regression model with the same input data. The prediction tests show that the ANN models used in this study have the potential of forecasting dust storm occurrence in Northwest China by using conventional meteorological variables. 相似文献
19.
《Physics and Chemistry of the Earth, Part C: Solar, Terrestrial & Planetary Science》2000,25(4):275-280
Multilayer feed-forward neural network models are developed to make three-hour predictions of the planetary magnetospheric Kp index. The input parameters for the networks are the Bz-component of the interplanetary magnetic field, the solar wind density n, and the solar wind velocity V, given as three-hour averages. The networks are trained with the error back-propagation algorithm on data sequences extracted from the 21st solar cycle. The result is a hybrid model consisting of two expert networks providing Kp predictions with an RMS error of 0.96 and a correlation of 0.76 in reference to the measured Kp values. This result can be compared with the linear correlation between V(t) and Kp(t + 3 hours) which is 0.47. The hybrid model is tested on geomagnetic storm events extracted from the 22nd solar cycle. The hybrid model is implemented and real time predictions of the planetary magnetospheric Kp index are available at http://www.astro.lu. se/-fredrikb. 相似文献
20.
Abstract— We report Rb-Sr analyses of phosphates from nine ordinary chondrites, more than doubling the number of meteorites for which such data are available. Ordinary chondrite phosphates characteristically have Rb/Sr ratios sufficiently low to permit accurate identification of initial 87Sr/86Sr, which is generally (but not in all cases) found to be significantly higher than the more primitive initial 87Sr/86Sr ratios inferred for carbonaceous chondrite refractory inclusions (ALL), basaltic achondrites (BABI), or bulk ordinary chondrites (in the ALL-BABI range). Such elevation of initial 87Sr/86Sr is generally considered to reflect isotopic redistribution during metamorphism, and with a model for Rb/Sr in this environment can lead to an inferred metamorphic timescale. For whole rock Rb/Sr the inferred formation intervals are typically tens of Ma (range nil to > 100 Ma). There is no evident relation between initial Sr elevation and metamorphic grade. There is not a clear difference in initial Sr effects between H and L chondrites; LL chondrites show much less (if any) elevation of initial Sr, but data are available for only two meteorites. For the first time it is possible to make a detailed comparison of initial Sr and I-Xe chronologies for several meteorites: these two potential metamorphic chronometers conspicuously fail to agree, in terms of both age and sequence of ages. A comparably definitive assessment of the comparison between initial Sr and Pb-Pb chronologies is not yet possible, but presently available data suggest that these two approaches to chondrite chronology also fail to agree. Without a correlation with metamorphic grade, or detailed agreement with an independent chronometer, it remains unclear whether initial 87Sr/86Sr in phosphates can be translated into a reliable chronometer for ordinary chondrite metamorphism, at least within simple interpretational frameworks. 相似文献