Undergraduate ALFALFA
Answers to scavenger1 questions
- Find Kornreich et. al. paper in AJ and explain how to do it:
I went into the Astronomical Data Base (ADS). Within ADS I selected "search references" and selected astrophysics journals ("Astronomy & Astrophysics"). I entered Kornreich and 2000 for the year of publication( Kornreich et. al. 2000, Astro J. 120,139.). Then I selected the journal article:"Departures from Axisymmetric Morphology and Dynamics in Spiral galaxies".
I could also have used the library on the second floor, where the astronomy journals are found.
- Figure out how to precess those J2000 coordinates to B1950:
I went into the NED database and in the tools table I choose "Coordinate transformation & Extinction Calculator". There, I put NGC 5701's B1950 coordinates from Kornreich's article: RA = 14h36m41.5s, dec = 05d34m50s. The calculator was set to give the J2000 coordinates which were: RA = 14h39m11.2s, dec =05d21m56.5s.
HI is detected in beams 1 and 6 at 14h39m11s:
beam J2000 B1950.0 1 14h39m11s 05d20m12s 14h36m41.3s 05d33m06s 6 14h39m11s 05d18m06s 14h36m41.3s 05d31m00s HI is visible in beam 6 from 14h38m32.7s to 14h39m56s:
beam J2000 B1950.0 6 14h38m32.7s 05d18m06s 14h36m03s 05d31m01s 6 14h39m56s 05d18m06s 14h37m26.3s 05d30m57s I checked the HI column density map, For Adrienne's first comment: HI from NGC 5701 is detected in beams 1 and 6. ALFA would have been located below the galaxy's visible image but centered in Right Ascension. We would expect HI detection in those beams.
For Adrienne's second comment HI is visible within a range in RA: The range in RA is wider than the HI map. From the map there is HI emission from 14h36m17s to 14h37m10s so we have detected an extended emission from this galaxy.
- Smoothing in the frequency domain:
- In the digital HI spectrum file the columns have names. Column 3, which is the raw HI data, is in Km/s and column 6, which is the final flux density, has units of mJy.
- Plot the raw spectrum:
I plotted the raw spectrum using an IDL program:
PRO PROFILE
; this program plots the raw spectrum
;the data is read from the file spectrum
;the velocity is stored in x and the flux density is stored in y
; then, a poscript file is created
openr,lun,'spectrum',/get_lun
data=fltarr(6,512)
readf,lun,data
x=data[1,*]; Segunda columna es velocidad
y=data[2,*];tercera columna es flujo
plot,x,y,title='NGC 5701 Raw',xtitle='velocity (Km/s)',ytitle='S (mJy)'
set_plot,'ps'
device,filename='NGC5701raw.ps'
plot,x,y,title='NGC 5701 Raw',xtitle='velocity (Km/s)',ytitle='S (mJy)'
device,/close
set_plot,'x'
END
- Three channel Hanning smoothing:
I did the three channel Hanning smoothing with the following program:
pro han
;este programa hace un hanning smoothing de tres canales.
;se lee el espectro de un file
;despues del smoothing hace un plot.
openr,lun,'spectrum',/get_lun
data=fltarr(6,512)
readf,lun,data
for i = 1,510 do data[2,i] = (0.25*data[2,i-1]+0.50*data[2,i]+0.50*data[2,i+1])
data[2,0]=data[2,1]
data[2,511]=data[2,510]
x= data[1,*];Segunda columna es velocidad
y=data[2,*];tercera columna es flujo pero con hanning
plot,x,y,title='NGC 5701 han',xtitle='velocity(Km/s)',ytitle='S(mJy)'
set_plot,'ps'
device,filename='NGC5701han.ps'
plot,x,y,title='NGC 5701 Han',xtitle='velocity(Km/s)',ytitle='S(mJy)'
device,/close
set_plot,'x'
end
- I did a three and five Boxcar smoothing with the following IDL programs:
pro box3
;este programa hace un boxcar smoothing de tres canales.
;se lee el espectro de un file
;despues del smoothing hace un plot.
openr,lun,'spectrum',/get_lun
data=fltarr(6,512)
readf,lun,data
for i = 1,510 do data[2,i] = (data[2,i-1]+data[2,i]+data[2,i+1])/3
data[2,0]=data[2,1]
data[2,511]=data[2,510]
x= data[1,*];Segunda columna es velocidad
y=data[2,*];tercerra columna es flujo pero boxcar smoothed 3
plot,x,y,title='NGC 5701 Boxcar 3',xtitle='velocity(Km/s)',ytitle='S(mJy)'
set_plot,'ps'
device,filename='NGC5701box3.ps'
plot,x,y,title='NGC 5701 Boxcar 3',xtitle='velocity(Km/s)',ytitle='S(mJy)'
device,/close
set_plot,'x'
end
Five channel Boxcar:
pro box5
;este programa hace un boxcar smoothing de cinco canales.
;se lee el espectro de un file
;despues del smoothing hace un plot.
openr,lun,'spectrum',/get_lun
data=fltarr(6,512)
readf,lun,data
for i = 2,509 do data[2,i] = (data[2,i-2]+data[2,i-1]+data[2,i]+data[2,i+1]+ data[2,i+2])/5
data[2,0]=data[2,2]
data[2,1]=data[2,2]
data[2,511]=data[2,509]
data[2,510]=data[2,509]
x= data[1,*];Segunda columna es velocidad
y=data[2,*];tercerra columna es flujo pero boxcar smoothed
plot,x,y,title='NGC 5701 Boxcar 5',xtitle='velocity(Km/s)',ytitle='S(mJy)'
set_plot,'ps'
device,filename='NGC5701box5.ps'
plot,x,y,title='NGC 5701 Boxcar 5',xtitle='velocity(Km/s)',ytitle='S(mJy)'
device,/close
set_plot,'x'
end
- Impact of smoothing on appearance of the profile: The three methods served the purpose of smoothing the raw graph.
- What is the difference in appearance among the different smoothings you tried? Boxcar 5 has the lowest noise; boxcar 3 and Hanning are very similar.
- Smoothing in the spatial domain:
- How long does it take a source to drift through the beam? It takes 14s for a source to drift through the beam considering the beam is 3.5' wide and the sky drifts by at 0.25 arcmin/sec.
- Explore the effect of smoothing by convolving an HI distribution with a Gaussian function.
I looked up the definition of convolution in the web at:
http://mathworld.wolfram.com/Convolution.html
I studied the CONVOL function in IDL. I also learned to do exponentiation with this ^ symbol, how to use the floating point value of !pi, and how to evaluate the Gaussian function at
http://mathworld.wolfram.com/GaussianFunction.html
I wrote two programs: one for the convolution with the triangle and another with the Exponential function with the raw spectrum.
This is the triangle convolution program:
pro convol5
;This program calculates the convolution of the raw data from NGC5701 with a One dimension triangle
;it first reads the file spectrum that is then placed in a matrix
;called data. The raw spectrum is read from the appropiate column in data,
;it then constructs a triangle function and it performs a convolution with it.
openr,lun,'spectrum',/get_lun
data=fltarr(6,512)
readf,lun,data
h1=make_array(512)
for i=0,511 do h1[i]=data[2,i]
openr,lun,'triangle',/get_lun
data=fltarr(2,49)
readf,lun,data
h2= make_array(49)
for i = 0,48 do h2[i]=data[1,i]
x1=convol(h1,h2)
x1 = x1/550
plot, h1, title= 'NGC 5701 RAW + 1-D triangle convolution',xtitle='velocity (Km/s)',ytitle= 'S (mJy)'
set_plot, 'ps'
device, filename = 'NGC5701triangle.ps'
plot, h1, title= 'NGC 5701 RAW + 1-D triangle convolution',xtitle='velocity (Km/s)',ytitle= 'S (mJy)'
oplot,x1
device,/close
set_plot, 'x'
end
This is the Program that convolves the raw data with the Gaussian Function.
PRO CONVOLUTION
; This program calculates the convolution of the raw data from NGC
; 5701 with a one Dimension Gaussian function
; it first reads the file spectrum that is placed in the matrix called
; data
; the raw spectrum of HI is read from the appropiate column in data
; it then constructs a Gaussian function that is centered in mu = 15
; and has a width (sigma)= 15.
; It performs the convolution and scales it with a factor of two.
openr, lun,'spectrum',/get_lun
data=fltarr(6,512)
readf, lun,data
h1=make_array(512)
for i = 0,511 do h1[i]= data[2,i]
x = make_array(25)
sigma = 15
mu = 15
const = 1/(sigma*sqrt(2*!pi))
for i = 0,24 do x[i]= data[0,i]
F= const*( EXP(-1.0*(x - mu)^2/(2*sigma^2)))
z = convol(h1,F,/center)
z = z*2
plot,h1, title='NGC 5701 RAW + Convolution',xtitle='velocity (Km/s)',ytitle='S(mJy)'
oplot, F
oplot,z
set_plot,'ps'
device, filename ='NGC5701convolution.ps'
plot,h1, title='NGC 5701 RAW + Convolution',xtitle='velocity (Km/s)',ytitle='S(mJy)'
oplot, F
oplot,z
device,/close
set_plot, 'x'
END
