Outline - Universidad Politécnica de Madrid
Contents
1. This willadd a second plasma model to the model defined at the beginning You will be asked to give initial values to the parameters ofthe new model Obviously the abundance should be the same than in the other thermal model To fix it when XSPEC ask you to give the initial value for the abundance type 3 apec Abundanc gt 7 N B parameter n can be fixed equal to parameter m at any time during the fitting process by typing newpar n m 2 Fit the spectra XSPEC gt fit Here we recommend making proofs with different initial values for each parameter and letting XSPEC fit the model to the spectra Ifthe results are not good for instance if the second temperature has no sense you can delete the model by typing XSPEC gt delco 2 apec 3 Plot the results XSPEC gt plot Idata del Then the new multi temperature model is over plotted to the spectra XSPEC plots the whole model wabs apec apec including both thermal components If you want to see the two thermal components separately you could type plot model To turn back to the fit type again plot Idata del XSPEC inspecting the results goodness of the fit Sometimes happens that small variations in the initial values produce different results when fitting the model with similar goodness of fit Here the goodness of fit is measured by the Chi Square statistics although other statistics could be used inside XSPEC The value of the chi sq2. of this source with SAS you will need to know its position in the image and its extension or better the extension of its PSF in the image For this exercise we will use only physical coordinates given by DS9 in the top panel For the extension ofthe PSF as an approximation you could use 330 pixels 17 arcsec N B You may need to change the zoom to better establish the centre of the source 6 Create the light curve with EVSELECT gt evselect table 0682_0148300101_EPN_S003_ImagingEvts ds withrateset Y rateset curve fits timebinsize 200 maketimecolumn Y makeratecolumn Y Y expression XMMEA_EP amp amp Pl in 300 5000 amp amp PATTERN lt 4 amp amp X Y IN circle xpoint ypoint 330 This will extract the light curve of a source situated in physical coordinates xpoint and ypoint remember to change xpoint and ypoint by the values obtained in DS9 in a radius of 330 pixels from the centre For this example we have chosen an energy PI range of 0 3 to 5 0 KeV You can also change it and extract different curves in different energy ranges The time bin size is 200 seconds 7 Inspect the light curve The output file curve fits can be opened with the task FV of HEASOFT Just type gt fv curve fits Two windows are opened the first one with the FV menu and the second one with some columns In the line with Extension name RATE mark plot in the view column Another window is
3. ASTROCAM School Young Stellar Objects from cool stars to exoplanets Eli mi AA 90 Timo gt Tartas DIM El Esconal Madrid 29 June 3 July 2009 If LETELIV Lab session 2 Coronal X ray emission general spectroscopic parameters and light curves Javier L pez Santiago amp Raquel M Martinez Arn iz Dpto Astrofisica Universidad Complutense de Madrid Spain Coronal X ray emission general parameters and light curves Introduction The final products of the reduction packages of the X ray missions XMM Newton SAS and Chandra CIAO are binary tables in fits format containing the events registered in each one of the instrument on board the satellite one table for each instrument Each line of the table of events corresponds to a single event and it offers some information such as the time at which the event occurred its position in the camera in which it was registered and its energy in electron volts eV Generally not all the events listed in the tables are good ones Here we are not going to discuss on how to reject bad events We encourage the reader to get the user manual of XMM and Chandra to see a detailed discussion on this issue We already selected the good events from the table of events in your home directory The reduction packages also allow the user to extract spectra and light curves of the sources detected in the observation However no other study can be
4. be fixed to a value N B after fixing or liberating a new parameter nothing is changed in the fit until you type again fit in the terminal 3 Finally to plot the results both model and residuals XSPEC gt plot Idata del XSPEC changing model parameters A useful thing in XSPEC is the possibility of changing the values of any parameter at any time you want This is made with the command newpar Contrarily to the case explained above when the value of a parameter is changed with newpar it immediately affects the model 1 Change the value of the temperature KT in the model XSPEC gt newpar 2 3 0 2 Now plot the data with the model again XSPEC gt plot Idata del Now in the graphic window you should see the effects of changing the temperature in the model The fit should not be good now If you type fit again in the terminal a new fit is performed and the new value of kT is given XSPEC adding new models In general stellar coronae are not well defined by a unique thermal model As in the case ofthe sun they present a gradient in temperatures In some cases generally with cool coronae one can obtain a good result when fitting a model with only one temperature 1 T model but usually multi temperature models give better results Only when the total number of counts in the X ray spectrum is small 1 T models produce good results 1 To add a second model type XSPEC gt addco 2 apec
5. carried out SAS and CIAO have no specific tasks to for instance study an extracted spectrum This needs to be done using specific existing software such as XSPEC or our own specifically created software The aim of this exercise is 1 to learn how to study some parameters of the corona of young stars by fitting hot plasma models to low resolution X ray spectra with XSPEC and 2 to obtain the light curve of at least a source detected in an XMM Newton observation in the proximities of a young cluster using specific tasks in SAS The following manual was created to be used as a tutorial Instead of reading everything before executing any action in the computer we suggest the student to read each point and perform the action one by one XSPEC fitting models 1 Start XSPEC by typing xspec11 in the terminal prompt you will need to load Heasoft of HEASARC first by typing head in the terminal if it was not done first gt head This is an alias created by us in your shell environment that invokes Heasoft If you have any problem just try gt source HEADAS headas init csh for cshell environments Immediately after type gt xspec11 Now the terminal prompt should be changed to XSPEC gt 2 Load the EPIC PN spectrum data XSPEC gt data n PN_binned_spectrum where n is the spectrum number By default n 1 So if n is not given the spectrum will take the value n 1 PN_binned_
6. ers APENDIX working with grppha If you want to prepare the spectra by yourself you will need to use the task GRPPHA available in the Ftools sub package of HEASOFT GRPPHA is an interactive command driven task to define or redefine and or display the grouping binning and quality flags Eventually you can append some information to the header of the resultant file as for instance the calibration files to be used The task can be used interactively or in a single command In the following we show two examples of using GRPPHA with a PN spectrum the same should be done for MOS spectra 1 Interactive use After typing grppha you are requested to give the name of the original spectrum e g src_spec pn fits Then the output name is requested in this example we will assume that the binning will be 6 so we will use the name extension pha6 input filename src_spec pn fits output filename src_spec pn pha6 The name extension is optional You could write another one but it is better to use extensions that will permit you to recognise the binning parameters easily Then you could optionally add some information to the output file such as the name of the background file and the calibration files to be used by XSPEC simply typing chkey BACKFILE bkg_spec pn fits chkey RESPFILE PN rmf chkey ANCRFILE PN arf Finally set the binning and exit group min 6 exit 2 GRPPHA in a single line gt
7. grppha infile src_spec pn fits outfile src_spec pn pha6 comm chkey BACKFILE bkg_spec pn fits amp chkey RESPFILE pn rmf amp chkey ANCRFILE pn arf amp group min 6 amp exit SAS generate light curves 1 Start SAS by typing sasrun in your home directory gt sasrun SASRUN is an alias created in your cshrc in your home directory It first loads the HEASOFT package and then run SAS the XMM Newton reduction software 2 Display the PN image in your home directory to look for X ray sources gt ds9 pn ds amp 3 Inspect the PN image To inspect the image you will need to use appropriate binning zoom and scale In the DS9 top menu go to bin and select block 32 this is a good choice for our image Now go to zoom and select zoom 1 2 The whole field of view should be now visible in the DS9 window Finally change the scale to logarithmic in the scale menu 4 Change colour map contrast and brightness optional The colour map the contrast and the brightness of the image can be also changed in DS9 The colour map is changed in the top menu Go to color and choose one of the maps we strongly recommend blackbody bb The contrast and brightness are changed simply moving the mouse while pressing the right button 5 Select the source to extract the light curve After step 4 you should be able to see several sources in the image Select one of them To extract the light curve
8. opened with some lines in the left panel and the plotting parameters in the right panel Now you only need to write TIME in the X box and RATE in the Y one and mark GO in the bottom panel The light curve should now be plotted in a different graphic window
9. spectively and n 3 for the PN data 7 Create the output graphic plotting window XSPEC gt cpd xs This will open a PGPLOT graphic window where the data and models will be plotted 8 Change to energy coordinates for plotting XSPEC gt setplot energy 9 Ignore bad channels and energy ranges where the matrix are not currently calibrated XSPEC gt ignore 0 0 0 3 for XMM Newton data XSPEC gt ignore bad N B if the energy range is given in float format then XSPEC understand that this is given in energy units Contrarily if the numbers are given in integer format without points e g ignore 0 1 then XSPEC understand that the range is given in channel units We strongly recommend using always energy units N B 2 Sometimes the calibration of the X ray spectrum at high energies is not very good for instance when the signal of the source is not very high or when if was a high background event during the observation that cannot be discarded In those occasions it could be a good idea to ignore also the data at hard energies For this exercise a good choice is ignore 5 0 that will ignore all the data with energy over 5 0 keV 10 Plot the data with X and Y axis in logarithm XSPEC gt plot Idata At this stage you should see your data three spectra plotted in the PGPLOT window White colour is used for the first spectrum n 1 in step 2 red for the second one n 2 and green for
10. spectrum is the name of the PN spectrum prepared with grppha see appendix Here you could use one of the spectra in your working directory with the extension pha prepared by us first e g src_spec pn pha50 3 Load the EPIC PN redistribution matrix omit this step if PN rmf was fixed during the use of grppha with chkey RESPFILE PN rmf XSPEC gt response n PN rmf Here n must take the same value than in step 2 Also in the next step step 4 n has to be the same number The calibration files redistribution matrix and ancillary file are different for each spectrum and chip MOS1 MOS2 or PN 4 Load the EPIC PN ancillary file omit this step if the pn arf was fixed during the use of grppha with chkey ANCRFILE PN arf XSPEC gt arf n PN arf 5 Load the EPIC PN background spectrum data omit this step if the background was fixed during the use of grppha with chkey BACKFILE background fits XSPEC gt backgrnd n background fits Where background fits is the name of the background spectrum This file is in the working directory with the name bkg_spec pn fits 6 Repeat steps 1 to 5 with the EPIC MOS1 and EPIC MOS2 data Remember to change the parameter n for each spectrum If you forgot to change it the first spectrum will be overwritten each time you use the command data You could use for instance n 1 and n 2 for the MOS1 and MOS2 data re
11. the third one n 3 XSPEC fitting data XSPEC fits automatically the model to each spectrum so all the commands given in the following are typed only once instead of three times For this exercise a hot plasma model with absorption will be used Here the absorption model is a multiplicative model that represents both the interstellar and circumstellar absorption 1 Load the model XSPEC gt model wabs apec Where WABS is the absorption model which has only one parameter the column density Nu and APEC is the hot plasma model with four different parameters temperature abundance redshift and normalization constant After loading the model you will be asked to give initial values to each variable You can use the default values given by XSPEC or change them Obviously the redshift should be set to zero in our stellar case 2 Fitting the model XSPEC gt fit 300 300 is the maximum number of iterations This will fit the model to our data showing at the end the results for each fitted parameter By default the abundance remains fixed in XSPEC when using APEC To liberate it you should write XSPEC gt thaw n Where n is the number of the parameter in the list of parameters given by XSPEC To know it type XSPEC gt show par One could also fix other parameters such as the temperature or the column density This is made by typing XSPEC gt freeze n Where n is the number of the parameter to
12. uared is given after the fitting process by XSPEC Good values of the reduced chi squared are around 1 XSPEC permits to study the parameter space Thus one could see iftwo parameters are correlated or if some parameters are not well defined 1 Use steppa to create stepped parameters from a given value see example bellow XSPEC gt steppa par1 low1 high1 num1 par2 low2 high2 num2 Here par1 and par2 are the list numbers of two different parameters of the model say abundance and column density low and high are the minimum and the maximum values of this parameter to be checked and num is the number of steps For example imagine you are using a model where the first parameter is the column density and the third one is the abundance Then the sequence XSPEC gt steppa 1 0 0 0 1 10 3 0 01 1 0 10 will perform fits to the data with the column density ranging from 0 0 to 0 1 10 steps and the abundance ranging from 0 01 to 1 10 steps The result is a matrix with the chi squared of each fit which can be plotted in the graphic device 2 Plot results of steppa XSPEC gt plot contour This will plot the results of steppa in the graphic window To turn back to the model plot type plot Idata del For this exercise we recommend to use steppa first with the column density and the abundance and then with the two temperatures of the 2 T model to see if one of these four parameters is less defined than the oth
Download Pdf Manuals
Related Search