An Active Flexure Compensation Method for LAMOST spectrograph based on BP-Neural Network

The Large sky Area Multi-Object fiber Spectroscopic Telescope(LAMOST)is a Chinese national scientific research facility operated by National Astronomical Observatories,Chinese Academy of Sciences(NAOC).The LAMOST survey for the Milky Way Galaxy and extra-galactic objects has been carried out for several years.The accuracies in measuring radial velocity are expected to be 5 km s-1 for the low resolution spectroscopic survey(R=1800),and 1 km s-1 for the medium resolution mode.The stability of spectrograph is the main factor affecting the accuracies in measuring radial velocity,so an Active Flexure Compensation Method(AFCM)based on Back Propagation Neural Network(BPNN)is proposed in this paper.It utilizes a deep BP(4-layer,5-layer etc.)model of thermal-induced flexure to periodically predict and apply flexure corrections by commanding the corresponding tilt and tip motions to the camera.The spectrograph camera system is adjusted so that the positions of these spots match those in a reference image.The simulated calibration of this compensation method analytically illustrates its performance on LAMOST spectrograph.

Gamma/hadron segregation for a ground based imaging atmospheric Cherenkov telescope using machine learning methods: Random Forest leads

A detailed case study of γ-hadron segregation for a ground based atmo- spheric Cherenkov telescope is presented. We have evaluated and compared various supervised machine learning methods such as the Random Forest method, Artificial Neural Network, Linear Discriminant method, Naive Bayes Classifiers, Support Vector Machines as well as the conventional dynamic supercut method by simulating triggering events with the Monte Carlo method and applied the results to a Cherenkov telescope. It is demonstrated that the Random Forest method is the most sensitive machine learning method for γ-hadron segregation.

Noise reduction methods in the analysis of near infrared lunar occultation light curves for high angular resolution measurements

A lunar occultation （LO） technique in the near-infrared （NIR） provides angular resolution down to milliarcseconds for an occulted source, even with ground- based 1 m class telescopes. LO observations are limited to brighter objects because they require a high signal-to-noise ratio （S/N ~40） for proper extraction of angular diameter values. Hence, methods to improve the S/N ratio by reducing noise using Fourier and wavelet transforms have been explored in this study. A sample of 54 NIR LO light curves observed with the IR camera at Mt Abu Observatory has been used. It is seen that both Fourier and wavelet methods have shown an improvement in S/N compared to the original data. However, the application of wavelet transforms causes a slight smoothing of the fringes and results in a higher value for angular diameter. Fourier transforms which reduce discrete noise frequencies do not distort the fringe. The Fourier transform method seems to be effective in improving the S/N, as well as improving the model fit, particularly in the fainter regime of our sample. These methods also provide a better model fit for brighter sources in some cases, though there may not be a significant imorovement in S/N.

Study of the continuum removal method for the Moon Mineralogy Mapper(M^3) and its application to Mare Humorum and Mare Nubium

The absorption band center of visible and near infrared reflectance spectra is a key spectral parameter for lunar mineralogical studies, especially for the mafic minerals（olivine and pyroxene） of mare basalts, which have two obvious absorption bands at 1000 nm（Band I） and 2000 nm（Band II）. Removal of the continuum from spectra, which was developed by Clark and Roush and used to isolate the particular absorption feature, is necessary to estimate this parameter. The Moon Mineralogy Mapper（M3） data are widely used for lunar mineral identification. However, M3 data show a residual thermal effect, which interferes with the continuum removal, and systematic differences exist among optical data taken during different optical periods. This study investigated a suitable continuum removal method and compared the difference between two sets of M3 data taken during different optical periods, Optical Period 1B（OP1B）and Optical Period 2A（OP2A）. Two programs for continuum removal are reported in this paper. Generally,a program respectively constructs two straight lines across Band I and Band II to remove the continuum,which is recommended for locating band centers, because it can find the same Band I center with different right endpoints. The optimal right endpoint for continuum removal is mainly dominated by two optical period data at approximately 2480 and 2560 nm for OP1 B and OP2 A data, respectively. The band center values derived from OP1 B data are smaller than those derived from OP2 A data in Band I but larger in Band II, especially for the spectra using longer right endpoints（〉2600 nm）. This may be due to the spectral slopes of OP1 B data being steeper than those of OP2 A data in Band I but gentler in Band II. These results were applied to Mare Humorum and Mare Nubium, and the measurements were found to mainly vary from intermediate- to high-Ca pyroxene.

Forecast daily indices of solar activity, F10.7, using support vector regression method

The 10.7 cm solar radio flux （F10.7）, the value of the solar radio emission flux density at a wavelength of 10.7 cm, is a useful index of solar activity as a proxy for solar extreme ultraviolet radiation. It is meaningful and important to predict F10.7 values accurately for both long-term （months-years） and short-term （days） forecasting, which are often used as inputs in space weather models. This study applies a novel neural network technique, support vector regression （SVR）, to forecasting daily values of F10.7. The aim of this study is to examine the feasibility of SVR in short-term F10.7 forecasting. The approach, based on SVR, reduces the dimension of feature space in the training process by using a kernel-based learning algorithm. Thus, the complexity of the calculation becomes lower and a small amount of training data will be sufficient. The time series of F10.7 from 2002 to 2006 are employed as the data sets. The performance of the approach is estimated by calculating the norm mean square error and mean absolute percentage error. It is shown that our approach can perform well by using fewer training data points than the traditional neural network.

Mapping the Milky Way with LAMOSTⅠ:method and overview

We present a statistical method to derive the stellar density profiles of the Milky Way from spectroscopic survey data, taking into account selection effects. We assume the selection function, which can be altered during observations and data reductions, of the spectroscopic survey is based on photometric colors and magnitude. Then the underlying selection function for a line-of-sight can be recovered well by comparing the distribution of the spectroscopic stars in a color-magnitude plane with that of the photometric dataset. Subsequently, the stellar density profile along a line-of-sight can be derived from the spectroscopically measured stellar density profile multiplied by the selection function. The method is validated using Galaxia mock data with two different selection functions. We demonstrate that the derived stellar density profiles reconstruct the true ones well not only for the full set of targets, but also for sub-populations selected from the full dataset. Finally, the method is applied to map the density pro- files for the Galactic disk and halo, using the LAMOST RGB stars. The Galactic disk extends to about R = 19 kpc, where the disk still contributes about 10% to the total stellar surface density. Beyond this radius, the disk smoothly transitions to the halo without any truncation, bending or breaking. Moreover, no over-density corresponding to the Monoceros ring is found in the Galactic anti-center direction. The disk shows moderate north-south asymmetry at radii larger than 12 kpc. On the other hand, the R-Z tomographic map directly shows that the stellar halo is substantially oblate within a Galactocentric radius of 20 kpc and gradually becomes nearly spherical beyond 30 kpc.

A method and results of color calibration for the Chang'e-3 terrain camera and panoramic camera

The terrain camera （TCAM） and panoramic camera （PCAM） are two of the major scientific payloads installed on the lander and rover of the Chang＇e 3 mission re- spectively. They both use a Bayer color filter array covering CMOS sensor to capture color images of the Moon＇s surface. RGB values of the original images are related to these two kinds of cameras. There is an obvious color difference compared with human visual perception. This paper follows standards published by the International Commission on Illumination to establish a color correction model, designs the ground calibration experiment and obtains the color correction coefficient. The image qual- ity has been significantly improved and there is no obvious color difference in the corrected images. Ground experimental results show that： （1） Compared with uncor- rected images, the average color difference of TCAM is 4.30, which has been reduced by 62.1%. （2） The average color differences of the left and right cameras in PCAM are 4.14 and 4.16, which have been reduced by 68.3% and 67.6% respectively.

Automated estimation of stellar fundamental parameters from low resolution spectra: the PLS method

PLS （Partial Least Squares regression） is introduced into an automatic estimation of fundamental stellar spectral parameters. It extracts the most correlative spectral component to the parameters （Teff, log g and [Fe/H]）, and sets up a linear regression function from spectra to the corresponding parameters. Considering the properties of stellar spectra and the PLS algorithm, we present a piecewise PLS regression method for estimation of stellar parameters, which is composed of one PLS model for Teff, and seven PLS models for log g and [Fe/H] estimation. Its performance is investigated by large experiments on flux calibrated spectra and continuum normalized spectra at different signal-to-noise ratios （SNRs） and resolutions. The results show that the piecewise PLS method is robust for spectra at the medium resolution of 0.23 nm. For low resolution 0.5 nm and 1 nm spectra, it achieves competitive results at higher SNR. Experiments using ELODIE spectra of 0.23 nm resolution illustrate that our piecewise PLS models trained with MILES spectra are efficient for O ～ G stars： for flux calibrated spectra, the systematic offsets are 3.8%, 0.14 dex, and -0.09 dex for Teff, log g and [Fe/H], with error scatters of 5.2%, 0.44 dex and 0.38 dex, respectively; for continuum normalized spectra, the systematic offsets are 3.8%, 0.12dex, and -0.13 dex for Teff, log g and [Fe/H], with error scatters of 5.2%, 0.49 dex and 0.41 dex, respectively. The PLS method is rapid, easy to use and does not rely as strongly on the tightness of a parameter grid of templates to reach high precision as Artificial Neural Networks or minimum distance methods do.

Stellar spectra association rule mining method based on the weighted frequent pattern tree

Effective extraction of data association rules can provide a reliable basis for classification of stellar spectra. The concept of stellar spectrum weighted itemsets and stellar spectrum weighted association rules are introduced, and the weight of a single property in the stellar spectrum is determined by information entropy. On that basis, a method is presented to mine the association rules of a stellar spectrum based on the weighted frequent pattern tree. Important properties of the spectral line are highlighted using this method. At the same time, the waveform of the whole spectrum is taken into account. The experimental results show that the data association rules of a stellar spectrum mined with this method are consistent with the main features of stellar spectral types.

A solar radio dynamic spectrograph with flexible temporal-spectral resolution

Observation and research on solar radio emission have unique scientific values in solar and space physics and related space weather forecasting applications, since the observed spectral structures may carry important information about energetic electrons and underlying physical mechanisms. In this study, we present the design of a novel dynamic spectrograph that has been installed at the Chashan Solar Radio Observatory operated by the Laboratory for Radio Technologies, Institute of Space Sciences at Shandong University. The spectrograph is characterized by real-time storage of digitized radio intensity data in the time domain and its capability to perform off-line spectral analysis of the radio spectra. The analog signals received via antennas and amplified with a low-noise amplifier are converted into digital data at a speed reaching up to 32 k data points per millisecond. The digital data are then saved into a high- speed electronic disk for further off-line spectral analysis. Using different word lengths （1-32k） and time cadences （5 ms-10 s） for off-line fast Fourier transform analysis, we can obtain the dynamic spectrum of a radio burst with different （user-defined） temporal （5 ms-10 s） and spectral （3 kHz-320kHz） resolutions. This enables great flexibility and convenience in data analysis of solar radio bursts, especially when some specific fine spectral structures are under study.

A method for calculating probability of collision between space objects

A method is developed to calculate probability of collision. Based on geometric features of space objects during the encounter, it is reasonable to separate the radial orbital motions from those in the cross section for most encounter events that occur in a near-circular orbit. Therefore, the probability of collision caused by differences in both altitude of the orbit in the radial direction and the probability of collision caused by differences in arrival time in the cross section are calculated. The net probability of collision is expressed as an explicit expression by multiplying the above two components. Numerical cases are applied to test this method by comparing the results with the general method. The results indicate that this method is valid for most encounter events that occur in near-circular orbits.

Combination of terrestrial reference frames based on space geodetic techniques in SHAO:methodology and main issues

Based on years of input from the four geodetic techniques （SLR, GPS, VLBI and DORIS）, the strategies of the combination were studied in SHAO to generate a new global terrestrial reference frame as the material realization of the ITRS defined in IERS Conventions. The main input includes the time series of weekly solutions （or fortnightly for SLR 1983-1993） of observational data for satellite techniques and session-wise normal equations for VLBI. The set of estimated unknowns includes 3- dimensional Cartesian coordinates at the reference epoch 2005.0 of the stations distributed globally and their rates as well as the time series of consistent Earth Orientation Parameters （EOPs） at the same epochs as the input. Besides the final solution, namely SOL-2, generated by using all the inputs before 2015.0 obtained from short-term observation processing, another reference solution, namely SOL- 1, was also computed by using the input before 2009.0 based on the same combination of procedures for the purpose of comparison with ITRF2008 and DTRF2008 and for evaluating the effect of the latest six more years of data on the combined results. The estimated accuracy of the x-component and y-component of the SOL- 1 TRF-origin was better than 0.1 mm at epoch 2005.0 and better than 0.3 mm yr- 1 in time evolution, either compared with ITRF2008 or DTRF2008. However, the z-component of the translation parameters from SOL-1 to ITRF2008 and DTRF2008 were 3.4 mm and -1.0 ram, respectively. It seems that the z-component of the SOL-1 TRF-origin was much closer to the one in DTRF2008 than the one in ITRF2008. The translation parameters from SOL-2 to ITRF2014 were 2.2, -1.8 and 0.9 mm in the x-, y- and z-components respectively with rates smaller than 0.4 mmyr-1. Similarly, the scale factor transformed from SOL-1 to DTRF2008 was much smaller than that to ITRF2008. The scale parameter from SOL-2 to ITRF2014 was -0.31 ppb with a rate lower than 0.01 ppb yr-1. The external precision （WRMS） compared with IERS EOP 08 C04 of the combined EOP series was smaller than 0.06 mas for the polar motions, smaller than 0.01 ms for the UT1-UTC and smaller than 0.02 ms for the LODs. The precision of the EOPs in SOL-2 was slightly higher than that of SOL-1.

Interference coupling analysis based on a hybrid method: application to a radio telescope system

Working in a way that passively receives electromagnetic radiation from a celestial body, a radio telescope can be easily disturbed by external radio frequency interference as well as electro- magnetic interference generated by electric and electronic components operating at the telescope site. A quantitative analysis of these interferences must be taken into account carefully for further electro- magnetic protection of the radio telescope. In this paper, based on electromagnetic topology theory, a hybrid method that combines the Baum-Liu-Tesche （BLT） equation and transfer function is proposed. In this method, the coupling path of the radio telescope is divided into strong coupling and weak cou- pling sub-paths, and the coupling intensity criterion is proposed by analyzing the conditions in which the BLT equation simplifies to a transfer function. According to the coupling intensity criterion, the topological model of a typical radio telescope system is established. The proposed method is used to solve the interference response of the radio telescope system by analyzing subsystems with different coupling modes separately and then integrating the responses of the subsystems as the response of the entire system. The validity of the proposed method is verified numerically. The results indicate that the proposed method, compared with the direct solving method, reduces the difficulty and improves the efficiency of interference prediction.

An optimal method for the power spectrum measurement

An aliasing effect brought up by mass assignment onto Fast Fourier Transformation （FFT） grids may bias measurement of the power spectrum of large scale structures. In this paper, based on the Beylkin＇s unequally spaced FFT technique, we propose a new precise method to extract the true power spectrum of a large discrete data set. We compare the traditional mass assignment schemes with the new method using the Daub6 and the 3rd-order B-spline scaling functions. Our measurement of Poisson samples and samples of N-body simulations shows that the B-spline scaling function is an optimal choice for mass assignment in the sense that （1） it has a compact support in real space and thus yields an efficient algorithm （2） without any extra corrections. The Fourier space behavior of the 3rd-order B-spline scaling function enables it to be able to accurately recover the true power spectrum with errors less than 5% up to k 〈 kN. It is expected that such a method can be applied to higher order statistics in Fourier space and will enable us to have a precision capture of the non-Gaussian features in the large scale structure of the universe.

The Coude Echelle Spectrograph for the Lijiang 1.8 m telescope

The Coude Echelle Spectrograph(CES)was originally mounted on Xinglong 2.16 m telescope in 1994.When the High Resolution Fiber-fed Spectrograph(HRS)was commissioned at Xinglong 2.16 m telescope,the red path of CES was moved to Lijiang 1.8 m telescope,following some redesign and reinstallation.The CES of Lijiang 1.8 m telescope has the spectral resolution(R=λ/?λ)ranging from 20000 to 50000 corresponding to different slit widths.With a 2 k×2 k PI CCD,CES has a wavelength coverage between 570 nm to 1030 nm.In particular,the resolution of 37000 at 0.5 mm slit corresponds to 1.3′′on the sky.The limiting magnitude is V=11.5 with the use of the tip-tilt feedback system,and the S/N is about 40 for 1 hour exposure at 600 nm(R=37000).During the installation of CES,the tip-tilt mirror technology had successfully fulfilled high precision guiding and tracking for high resolution spectral observation and significantly improved the observation efficiency.

Investigating the efficiency of the Beijing Faint Object Spectrograph and Camera(BFOSC)of the Xinglong 2.16-m reflector

The Beijing Faint Object Spectrograph and Camera （BFOSC） is one of the most important instruments operating in conjunction with the 2.16-m telescope at Xinglong Observatory. Every year there are - 20 SCI-papers published based on observational data acquired with this telescope. In this work, we have systemically measured the total efficiency of the BFOSC that operates as part of the 2.16-m reflector, based on observations of two ESO flux standard stars. We have obtained the total efficiencies of the BFOSC instrument of different grisms with various slit widths in almost all ranges, and analyzed factors which effect the efficiency of this telescope and spectrograph. For astronomical observers, the result will be useful for them to select a suitable slit width, depending on their scientific goals and weather conditions during observations. For technicians, the result will help them to systemically identify the real efficiency of the telescope and spectrograph, and to further improve the total efficiency and observing capacity of the telescope technically.

The tidal tails of globular cluster Palomar 5 based on the neural networks method

The sixth Data Release （DR6） of the Sloan Digital Sky Survey （SDSS） provides more photometric regions, new features and more accurate data around globular cluster Palomar 5. A new method, Back Propagation Neural Network （BPNN）, is used to estimate the cluster membership probability in order to detect its tidal tails. Cluster and field stars, used for training the networks, are extracted over a 40 × 20 deg^2 field by color-magnitude diagrams （CMDs）. The best BPNNs with two hidden layers and a Levenberg-Marquardt （LM） training algorithm are determined by the chosen cluster and field samples. The membership probabilities of stars in the whole field are obtained with the BPNNs, and contour maps of the probability distribution show that a tail extends .5.42° to the north of the cluster and another tail extends 3.77° to the south. The tails are similar to those detected by Odenkirchen et al., but no more debris from the cluster is found to the northeast in the sky. The radial density profiles are investigated both along the tails and near the cluster center. Quite a few substructures are discovered in the tails. The number density profile of the cluster is fitted with the King model and the tidal radius is determined as 14.28＇. However, the King model cannot fit the observed profile at the outer regions （R 〉 8＇） because of the tidal tails generated by the tidal force. Luminosity functions of the cluster and the tidal tails are calculated, which confirm that the tails originate from Palomar 5.

A sodium laser guide star coupling efficiency measurement method

A large telescope＇s adaptive optics （AO） system requires one or more bright artificial laser guide stars to improve its sky coverage. The recent advent of a high power sodium laser is perfect for such application. However, besides the output power, other parameters of the laser also have a significant impact on the brightness of the generated sodium laser guide star, mostly in non-linear relationships. When tuning and optimizing these parameters it is necessary to tune based on a laser guide star generation performance metric. Although return photon fluxis widely used, variability of the atmosphere and sodium layer makes it difficult to compare results from different sites or even within a short time period for the same site. A new metric, coupling efficiency, is adopted in our field tests. In this paper, we will introduce our method for measuring the coupling efficiency of a 20W class pulse sodium laser for AO application during field tests that were conducted during 2013-2015.

The Coudé Echelle Spectrograph for the Xinglong 2.16m Telescope

A brief description of the NAO coude echelle spectrograph mounted on the 2.16 m telescope at Xinglong station is given. This echelle spectrograph is located at the coudé focus with a prism cross disperser. The echelle image covers the spectral region from 330 to 1100 nm displayed in 80 spectral orders in one exposure through two light beams. With a slit height of 2mm, spectral orders are separated by 15 to 23 pixels in blue region and by 7 to 19 pixels in red region. Alternatively, two additional resolution modes corresponding to different focal length cameras with resolving power R 16000, 170 000 in the blue beam and R = 13000, 170 000 in the red beam could be provided by this spectrograph. The bias, dark, wavelength calibration, flat field and science exposure are described in details. The limiting magnitude for 1 hour exposure with an S/N ratio of 100 scales to V = 9.5 in the red path and to V = 7.2 in the blue path.

Light curve inversion of asteroid(585)Bilkis with Lommel-Seeliger ellipsoid method

The basic physical parameters of asteroids, such as spin parameters, shape and scattering parameters, can provide us with information on the formation and evolution of both the asteroids themselves and the entire solar system. In a majority of asteroids, the disk-integrated photometry measurement constitutes the primary source of the above knowledge. In the present paper, newly observed photometric data and existing data on（585） Bilkis are analyzed based on a Lommel-Seeliger ellipsoid model. With a Markov chain Monte Carlo（MCMC） method, we have determined the spin parameters（period, pole orientation）and shape（b/a, c/a） of（585） Bilkis and their uncertainties. As a result, we obtained a rotational period of 8.5738209 h with an uncertainty of 9×10^-7h, and derived a pole of（136.46°, 29.0°） in the ecliptic frame of J2000.0 with uncertainties of 0.67°and 1.1°in longitude and latitude respectively. We also derived triaxial ratios b/a and c/a of（585） Bilkis as 0.736 and 0.70 with uncertainties of 0.003 and 0.03 respectively.