Chapter 11
Using SPLAT

Splat is a graphical spectral-analysis tool. Splat offers tools for further processing, fitting, or identification of spectral lines. One can also plot different spectra in the same window and make publishable files. The full Splat documentation can be found in SUN/243.

11.1 Opening a spectrum in SPLAT

There are two options for opening up a spectrum in Splat. First it is possible to send a spectrum to Splat using Gaia as described in Section 10.8.

Alternatively spectra can be displayed in directly in Splat using:

for a single position cube or

to display the spectrum at the pixel position (xpos,ypos). Or simply initialise Splat and open the file directly from the gui.

When Splat is initially started you will get a main window appear and also a spectral window.

Figure 11.1: The main Splat window and spectral window.

11.2 Display synopsis of spectrum

When you initially open a specrum in Splat the synopsis of that spectrum will automatically be overlaid on the spectrum displayed, if the metadata are known to Splat. The following are a list of possible properties than can be displayed, with a brief description. Names in parenthesis are either the FITS keywords or AST attributes used to obtain the values.

(1)

Name: the short name

(2)

Telescope: telescope, instrument and its backend (TELESCOP/INSTRUME/BACKEND)

(3)

Object: target, molecule and molecular transition being observed (OBJECT/MOLECULE/TRANSITI)

(4)

Date obs: date of the observation (DATE-OBS or Epoch in UTC)

(5)

Elevation: observatory elevation (ELSTART or $0.5\ast$(ELSTART$+$ELEND))

(6)

Exposure: exposure time for the spectrum (EXTIME)

(7)

Exposure (median): exposure time for the spectrum (EXP_TIME)

(8)

Exposure (elapsed): the time elapsed during exposure (INT_TIME otherwise DATE-END$-$ DATE-OBS)

(9)

Exposure (effective): effective exposure time (EXEFFT)

(10)

Coords sys: the co-ordinate system code (System)

(11)

Spec position: position of spectrum on the sky (EXRAX, EXDECX)

(12)

Src position: position of the observation centre (EXRRA, EXRDEC)

(13)

Img centre: centre of the originating image (EXRA, EXDEC)

(14)

Offset: offset of spectrum from the observation centre (EXRRAOF, EXRDECOF or EXRAOF, EXDECOF)

(15)

Doppler RA Dec: reference position (RefRA,RefDec)

(16)

SourceVel: source velocity (SourceVel)

(17)

SourceVRF: source velocity reference frame (SourceVRF)

(18)

SourceSys: system of the source velocity (SourceSys)

(19)

StdOfRest: standard of rest (StdOfRest)

(20)

RestFreq: rest frequency (RestFreq

(21)

ImageFreq: image sideband equivalent of the rest frequency (ImagFreq)

(22)

Channel spacing: channel spacing (derived value)

(23)

Number of channels: Number of co-ordinate positions

(24)

TSYS: system temperature (TSYS)

(25)

TSYS (median): median system temperature (MEDTSYS)

(26)

TSYS (est): system temperature (derived value)

(27)

TRX: receiver temperature (TRX)

To show or remove the synopsis of a spectrum simply click on the Options button in the spectral view window and select or deselect Display Synopsis. To move the position of the synopsis simply click and drag the text box to the desired location.

Figure 11.2: How to display or remove the synopsis of a spectrum in the spectral window.

11.3 Changing units of a spectrum in SPLAT

Sometime you may wish to change the units of a spectrum in Splat.

(1)

Click on Change the units of the current spectrum button in your current spectral window.

(2)

A new window will appear called "Current Spectrum Units".

---

Figure 11.3: Top: Select the Change the units of the current spectrum button. Note that in this example the current units are Frequency. Bottom: the "Current Spectrum Units" window.

---

(3)

From the Coordinates drop-down menu select the desired unit (in this case we select Kilometers-per-sec (radio)). Then click Apply.

---

Figure 11.4: Changing the units of the spectrum via the drop-down menu to km s$-1$.

---

(4)

You will now see that the spectrum is displayed in Radio Velocity (km/s).

---

Figure 11.5: Spectrum with units of km s$-1$.

---

11.4 Cropping a spectrum in SPLAT

(1)

Click on Cut out the regions of the current spectrum button above the spectrum you wish to work on. This will open a new window.

(2)

Select Add from the cut region window.

---

Figure 11.6: Top: Select the cut button. Bottom: Select the Add button.

---

(3)

Click and drag the regions you wish to remove from the spectrum in the spectrum window. As we had a second region to remove we click on the Add button a second time. In this example we select the two noisy edge regions from the spectrum to remove.

(4)

Select the Remove Selected button.

---

Figure 11.7: Top: The two regions selected. Bottom: remove the selected region by clicking the Remove Selected button.

---

(5)

A new spectrum with the noisy edges removed will now appear in the main Splat window. Click on your newly created spectrum to view.

---

Figure 11.8: Top: Select the newly created spectrum in the main Splat window. The spectrum with the selected noisy-edge-data removed will then appear.

---

11.5 Rebinning a spectrum in SPLAT

(1)

Click on Apply a filter to the current spectrum button above the spectrum you wish to work on. This will open a new window.

(2)

Select the number of channels over which you want to bin—this is the Width. In this example the spectrum has a resolution of 0.055 km s$-1$. We bin by 18 channels to rebin the spectrum to 1 km s$-1$ channels.

(3)

Select Filter (Replace). This will create a new spectrum in the main Splat window and will also replace the existing spectrum displayed in the spectral window with the new binned spectrum.

---

Figure 11.9: Top: The Splat "Filter regions of a spectrum" window. Bottom: The final rebinned spectrum.

---

11.6 Estimating the noise in a spectrum using SPLAT

(1)

Click on Get statistics on region of spectrum button above the spectrum you wish to work on. This will open a new window.

(2)

Select Add from the "Region statistics" window.

---

Figure 11.10: Top: Select the statistics button. Bottom: Select the Add button.

---

(3)

Click and drag the regions you calculate the statistics over. As we wish to include baselines eithersie of the spectral line seen in the data we click on the Add button a second time.

---

Figure 11.11: Click and drag the regions over which you wish to estimate the noise.

---

(4)

You will see the two selected spectral regions appear in a table in the "Region statistics" window with some basic statistic already computed for each region created.

---

Figure 11.12: Two spectral regions have been selected and are displayed in the "Region statistics" window.

---

(5)

Click on Selected stats or All stats to get statistics of the regions combined. In this example we see the standard deviation in the regions selected is 0.19K.

---

Figure 11.13: Statistics for both regions are displayed in the "Region statistics" window.

---

11.7 Fitting a line in a spectrum using SPLAT

Your data might have lines for which you want a quick estimate of its line strength and width.

(1)

Click on Fit spectral lines using a variety of functions button above the spectrum you wish to work on. This will open a new window.

(2)

Click on the Add button in the "Measure spectral lines" window.

---

Figure 11.14: Top: Select the spectral-fitting button. Bottom: Click on the Add button.

---

(3)

Click and drag the box that appears in the spectral window to encompass the line for which you wish to calculate a fit for.

---

Figure 11.15: Click and drag the region containing the line which you wish to be fitted.

---

(4)

Click on the Fit button to produce both a Quick and Gaussian fit to the spectral line.

---

Figure 11.16: In the "Measure spectral lines" window, we see the both the Quick fit (black dashed line) and the Gaussian fit (pink line) of the spectral line.

---

(5)

We see the results of the fit are provided in the "Measure spectral lines" window, and the two fits can be seen in the spectral line window.

---

Figure 11.17: In the spectral window we see the both the Quick fit (black dashed line) and the Gaussian fit (pink line) of the spectral line.

---